# Lessons in writing Questions

There are other posts on offering advice for writing questions, such as this list of suggestions. This post is different: it will give detailed lessons worked out on complete texts, showing before and after versions. Ideally, they will be based on real examples, not made up might-have-been examples thought up to show a point.

• I think adding direct links to each answer (lesson) within the question body would be a nice addition. – NVZ Aug 24 '17 at 13:42

# Lesson 2

A newcomer posted a question. The original was quickly put on hold as “too broad” but had other problems too.

Besides being a list of questions with no elaboration on any of them, asking for things like the mass of Ceres and the orbit diameters will get you downvoted, as that’s basic information shown in the infobox on the Wikipedia page. Furthermore, asking how much force is required means you don’t understand that you can trade force for time. Asking which propulsion system to use doesn’t affect the energy needed unless you mean total including carting the fuel. And that’s too broad and opinion based without some context.

A note about a question that can be asked as a single line without elaboration: if it’s asking for a clear fact, like “What is the mass of Ceres?” then it shows that the asker didn’t do basic research, which we consider rather rude. If it’s not a basic fact than it’s not self-contained in a single sentence and needs some context.

A later iteration was:

### Question: How much energy is required to move Ceres into orbit around Mars?

How could you move Ceres into orbit around Mars?

The Title and the Body are asking different things, and the single-sentence body lacks context, so it’s either “too broad” or “primarily opinion based” or both. Focusing on the question in the Body (leaving the old Title might have been an oversight), you would need to specify the civilization and tech level, etc. Is this a near future in our real world?

As an exercise, which I now share with you, I went about fleshing it out.

First and foremost, what is the nature of the civilization doing this? A super-advanced civilization with beyond-real-universe physics might just teleport it or order a suitibly sized FTL drive from a contractor, or find that bigger is easier using a spindizzy device.

Now I don’t know what context the original asker had in mind, but I can consider what kind of context would be good for discussing here. I chose near future with real-world science, specifically making it a stretch to have the civilization acheive this. This way the kinds of systems people are dreaming up now will be applicable to the answer, and it will be approachable and understandable to the reader.

Starting with that, my first attempt was posted in the Sandbox. This takes the style of giving story/setting/plot exposition as a means of explaining context. This is not always a sucessful template, as it tends to get long and dwell on unnecessary details. So the best thing is to keep it short.

Who are these people? This is needed context.

Why are they doing this? The details are far less necessary, but addresses a couple issues. Readers will wonder about certain arbitrary choices and restrictions, and explaining the movite or situation which generates these restrictions, rather than listing them in the abstract, will actually be easier and prevent answers that overlooked the points. So the trick here, especially, is to keep it to an executive summary level of detail. The “story flair” can be a mental picture that saves a laundry list of specifications, rather than being a superflorous ornament.

By mentioning how it is a political move, it is evocative of the U.S. Apollo missions. It does not need to make an immediate profit, but it ends up becoming a long-term investment in the society’s next steps. Perhaps some Answer will muse on this, but I don’t need to go over it here.

Feedback in the comments led to my adding another paragraph. This is what would happen in regular Question posts, too. Some newcomers will answer comment questions by posting another comment! That is not right! The comments are not a discussion forum, but (ideally) review notes. So address the comments by improving the post. It’s best to integrate the improvement into the post. It’s poorer to add clarifying notes as an appendix.

Finally, although the question is “ok” as it stands, it’s not “best”. It’s short enough that it can lead up to the actual question, with the Title only establishing where this is heading. But why not add a further lead in? Expanding on the Title question, I can put a question first: “What technology will humanity use the first time it moves a planet?”. This also focuses on choosing the technology as being the question, so the paragraph I added in response to comments might not be as necessary, or at least not so in-your-face.

The final text is posted as a question here.

# Results

There was no feedback concerning the usual post-closing issues, and nobody flagged the question for closing. Thus, the lesson is sound. As I review this, there are 14 upvotes (no downvotes) and two “stars”.

In fact, the question made it onto the Hot Network Questions list a short time after I posted it.

That brings me to the subject of titles, which is an issue I had in composing the question but left out of the narration above.

# Lesson on Question Titles

In order to get a lot of views, especially from cross-site listing under Hot Network, people choose to follow the link based only on the Title. In order to evaluate the writing of the body, they have already viewed it!

The best Title is a brief summary of the question, and that commonly implies that it is written as an interogative sentence. This advice is generally given when “bad” titles are seen. These are statements (not questions) that are overly general or generic, or even oblique to the real question contained within.

But I had trouble phrasing a question that was simpler than the opening sentence of the post’s body, and went with a phrase that’s not even a sentence.

The real point is not whether the Title is phrased as a question, but rather that it’s specific. There are bad titles that are too generic that are still questions, along the lines of “Will my plan work?” In my case, the implied how? is indeed a good indication of the question, but I was able to leave off a bunch of awkward connecting words as well as the interogative pro-form (although in 5th grade English they just taught it as being an interogative pronoun).

So although I agonized over how to make the title follow the best practice and decided against it, this worked and worked well. It attracted people, and was not “click bait” because it was accurate and indicitive of the full question (which is stated as the very first line in the body).

# Lesson 1

The general template we see here is that of a large dump of information, which may be consistent with the story plot, background, backstory, or details of a high concept; followed by a question that involves that. Typically the question is set off, and may be labeled “…and here’s the actual question:”.

Even in cases where the info is exactly relevant to the question, that's still an inferior template. The question should come first, and then needed information. Take a “spiral approach” if needed — I’ll post an example of that eventually.

Note that the feedback comments on this post include:

In one sentence, what is your question?

So my advice (in chat) started:

Rip it all out. Put that first, without needing to say “oh, here's my real question”. Just start by asking that. Then, follow up with some detail on the chemical in question.

Another experienced user, sphennings, had been presenting the following lesson:

2. provide (only) what’s needed to explain and give needed context.

So, I continued this plan, working the question text out from scratch.

Start with the question: “How can (these people) make hazmat suits for (some chemical)?”

Just trying to formulate the succinct question, you see that (these people) need to be described. This is, in fact, key to making the question not be “too broad”. This is not addressed in the original question at all, but it mentioned people on horseback. A modern society would have no issues making suitable gear and it’s a non problem. So I guessed — to make an interesting choice — Bronze Age technology.

So the question becomes: “How can Bronze Age people make hazmat gear for (some chemical)?”

Now the original post is nothing but details on a specific chemical, which is just too much irrelevant information. But, is the specific chemical even needed? Consider if the answer is “no” and it’s just some caustic or reactive chemical. Then, the answer would not be possible, as the author can just make up anything. He can proclaim that some specific material is not affected, or some specific treatment does the trick. That might be just the ticket for the plot, but it makes this a non-question. So, I'll stick with the specific chemical.

The reader will wonder where this volatile chemical comes from, without modern tech level. So you add “the chemical is produced biologically, just as fruits produce acid or capsaicin …they breed these into weapons that produce a quantity of concentrated chemical”

Having the stuff simply grow naturally fails a reality check. So this got revised into natural feedstocks that allows the people to produce it in a suitable simple matter.

Note that the high vs low tech problem is needed context. Other issues will be raised via comments. But don’t add gobs of exposition and backstory thinking that context is good just because you’re dumping more information.

We also need some description of the chemical, besides a link to Wikipedia. But keep it a short summary and include relevant details such as its problem with the obvious materials such as leather.

Working it through, I wind up with this, which I've posted as a new Question. And we shall see if it’s any better received than the original.

# Results

I posted the question at what (according to site analytics and my own experience) is the slowest time in WB, in terms of page views as well as activity. Though not especially slow as far as weekends typically get.

Even so, the question got 2 upvotes within 15 minutes of being posted, before I revealed this meta post. And my questions are not automatically well-received just because it’s me. So, I must conclude that the lesson is sound.

## review

After a few days, I see that there was still no criticism concerning the question being too broad, unclear what is being asked, primarily opinion based, or too story based, which are the main reasons Questions are closed. Nobody flagged it for closing, for any reason.

On the other hand, the voting was strange. It is a fairly even split of upvotes to downvotes. As I write this, it’s +14/−8, but for the first day or two after posting the total stayed around +2, with new votes evenly split. So, people apparently don’t like the question. I assume this has to do with the subject of the question, not the quality of the writing. The comments indicate that some people found the premise unbelievable or supposed it was a bad question for being unanswerable.

This is a good example of why you should not dismiss a question that you think has no solution! Will’s answer is well worth having asked the question.

• Sound, but not free from error. Still, good framework/analysis. – Frostfyre May 7 '17 at 1:39
• This is all good advice but a question can still tick all these boxes and not be well received. The question you posted was a sort of impossible ask. Yeah some people came up with ideas but if you want it science based the answer will always be one about reducing the number of deaths. I think perhaps your question assumed too much and you should start off with a "Could Bronze age people protect themselves from chlorine trifluoride" reality check and then work from there. I'm not sure 2 up votes is enough to conclude anything though. Not saying this post is wrong, just some thoughts :) – FreeElk May 7 '17 at 13:38
• @FreeElk I'll add another coda later. – JDługosz May 7 '17 at 13:41
• I think something that can be learned from this question is that unnecessary details can bog down the answering process. The details about bioavailable precursors to CIF3 aren't relevant to the question and seemed to have a lot of people confused. The question probably would have better just asking about protective gear. – sphennings May 8 '17 at 13:46
• @sphennings I'm a fan of parsimony in introducing universe changes, so I think it's elegant if the same thing that causes the (main thing) provides a solution, too. I was thinking of Erin Thursby’s reply for example, but foreshadowed that Teflon wouldn’t do it. Having the stuff magically appear would not provide the same avenues. – JDługosz May 8 '17 at 14:45
• @JDługosz Erin's answer is along the lines of "You handwaved one thing why not handwave something else?" Answers like that generally aren't received well by the community. – sphennings May 8 '17 at 14:52
• Yes, I would prefer something specific. But that's why I included the source as a main point in tne question. «You might suppose that the availability of fluorine-bearing molecules in the local biology will offer a solution, but the obvious Teflon is also ignited by the stuff» – JDługosz May 8 '17 at 15:01
• @sphennings I think Erin's answer is more along the lines of "You've handwaved something into your universe" (the plants able to cope with the material) "why not use that thing you've already created to do this job?" (unless I'm missing something). Less about handwaving two things, more about using the thing you did handwave already. – FreeElk May 14 '17 at 12:59

# Lesson 4 — Abstraction

When you are creating something — whether it be a science fiction or fantasy story in the conventional sense, a role-playing game scenario, or anything that I’ll lump together under unconventional fiction for now, you are deep into creating a scenario or describing some piece of technology when you run into a question.

The first thing that comes to mind, then, is to brain-dump what you were working on and append “I have a problem here.” This often has a breakhead The Actual Question near the bottom. This was seen in Lesson 1, for example. I’ve already explained how the narrative form of the question post is backwards and flawed in other ways. But now I’ll look at the deeper process that leads to this and how to create a good Question starting from this point.

When the question comes to mind, you are naturally steeped in the context and backstory of the thing you were really working on. The question came from that. So, you might think that the question necessarily involves all that stuff.

The key is to take a step back and abstract the issue in the question. I see two examples in the Sandbox’s recent activity.

# Total Recoil

In a draft question concerning how to handle recoil from ultra-powerful hand guns, the cure is not to add complete details of the project in which the issue arose. The underlying question, “How are you going to protect your hands from the recoil?” is completely general, and really ought to be something that’s considered before designing such powerful weapons.

So I’ll turn the question around and rather than asking how to handle the kick of some specific weapon, ask instead how the recoil issue may be mitigated in general and how much kick can the wielder handle given the proposed solutions.

### The lede

How can you handle the recoil of ultra-powerful sidearms?

What else do we need to know to fully understand the question from that one sentence? Well, the fact that it uses a railgun propulsion system is critical, since this inherently transfers momentum to the launcher, and can’t vent like a rocket launcher. Rather than expanding on “ultra-powerful sidearms” by complicating the sentence structure, this would be noted in another sentence. But using “railgun” instead of “sidearm” doesn’t not make the sentence more complex. On the other hand, it no longer implies the hand-held portable nature, and “hand-held railgun” does add complexity and is not as clean.

We also need to deal with the general tech level, but that’s easily put off to the Brief Context. Can the Synopsis simply end there and let the Brief Context carry it? The line between them, as we’ve seen before (Branching Timeline example) is fluid. So, I’ll skip the rest of Step 2 (Synopsis) and work on the Brief Context.

In the near future, advances in warm superconductors and battery technology has led to railgun projectile launching devices of enormous power. The energy storage (battery or fuel cell) can be carried separately and connected via a cord, but the superconductive rail that actually does the launching is hand-held and aimed like a small sidearm. The amount of energy given and the weight of the projectile can be varied, even on a shot-by-shot basis. The system is limited not by how powerful the gun can be, but by how much power the wielder can handle! Note that unlike a rocket launcher, the nature of the system demands that the recoil is felt directly by the gun itself.

Suppose the near-future technology has advanced to the point indicated in order to build the weapons described, plus perhaps a little more. Other areas of science will have advanced as well, and you can be flexible on this, as long as it seems like roughly the same time frame needed for the proposed advances.

In what ways might a soldier be able to cope with such a powerful weapon’s recoil? What is the effective limit to how much power he could use in a weapon that one person carries and easily points?

Now, I’m also thinking of possible directions the answers might take. I want to leave the question open enough to accommodate them, but restrict things so the answers don’t get too off track or irrelevant for what I have in mind.

If you include this idea list in the question post explicitly, it makes people wonder why you asked if you already had an answer in mind. If you have an idea of your own, you might not want to influence other readers until after they’ve had a chance to post. So list things to exclude, and be prepared to discuss them in the comments, but don’t give a list of possible answers to be elaborated upon.

I posted this question Here, and decided not to add more to the Synopsis paragraph. But, I used “railgun” in the Title.

Now, someone who came up with an issue designing a scenario might have no clue how to approach an answer. Since this is second hand and has already been discussed, I do have some ideas. I might need to “hint” or give more guidance if the question does not receive a favorable reception, or nobody covers some of the ideas I had. A question post can be supplemented with comments on the question and discussion in the chat room. It is revised over time. It can evolve as part of the community process of exploring it.

After you post a question, keep close to WB.SE so you can keep an eye on it! It’s rare to see a question post that has not had revisions, and it makes me wonder if the OP has gone away, especially if there are comments.

# results

The immediate results surprised me. Not only did I post at what is generally the slowest time of the week, the site analyitics (visible to those with 25k rep, IIRC) indicate that it’s a significant lul, a record low in page views since April 15 which was about the same. That the question immediatly got 5 answers posted in an hour accounts for 1/6 of the posts for the day! It made it onto the Hot Network Questions list.

Come monday, when traffic picked up, this question exceeded 2500 views, reached 10 upvotes, and 3 bookmark stars. More telling than the votes is the number of comments, both on the original post and on answers. Guns make for popular questions!

The question got several upvotes, and there were no comments asking for clarification or more details. There were no flags concerning the usual Question review queue issues.

This not only confirms that this lesson is sound in terms of writing, but it shows that the original sandbox poster had the germ of an interesting question, if only it could be abstracted from the original scenario in which the question arose.

As I pointed out before, I turned the question around from what was originally posted. This means that the results should feed back into the weapons design. This suggests an iterative design process for whatever is being written here — never be afraid to throw away what you’ve done if you realize you designed (or plotted) yourself into a corner!

# Lesson 3 — The Spiral Template and Inverted Pyramid

In the previous two lessons, I (re)developed a question post text starting with the question itself, and following from that. Now I’d like to contrast that with the way we’re taught to write essays in English Composition class, and also point out that this follows a spiral approach.

In both high school and college here in the United States, I was taught certain forms and conventions for writing essays. Later I taught the same.

To write an essay, the thesis statement should be the last sentence in the first paragraph. This contrasts to the template I’m promoting here, which is where you put a simple expression of the question as the very first sentence. There are essay forms for comparison and contrast, explaining how to do something, persuading a point of view, and a few others; but none for asking a question.

So, the template (general layout and rules to follow) are dissimilar to what you probably learned in Writing class. It’s more akin to the way a classic newspaper article was written. That’s why I adopt the terminology from journalism writing, and call that first important sentence the lede, rather than a thesis or a topic.

Here is an overview of the template I’ve been using in the first two lessons:

1. Title of post.
2. The lede — a single sentence (as a question) comes first. It should not be too complex, but as such it does not stand alone. It needs more details to fully understand. Keep it as a simple sentence! You don’t cram details in as subordinate clauses; rather, you gloss over anything you can and give the details in subsequent sentences, as noted in the next step.
3. The synopsis. This expands the single-sentence lede into a paragraph that gives just enough detail to grasp the meaning.
4. Brief Context. Paragraphs that explain the necessary details and background. This is still kept short — no more than one paragraph per important point, even if the discussion is not complete. The idea is to cover the entire subject horizontally, touching on each important point to some degree. This is as opposed to covering one point in exhaustive depth before going on to the next.
5. Restate the question, this time knowing that all the previous explanation is known.
6. Deep Context. More details and background that was left out of number 4.
7. Appendices and Codas — notes that don’t fit properly in the narrative structure.

Now earlier I referred to needing to add Deep Context as using a spiral approach. But really, it’s all spiraling! Points 1 through 5 are all distinct turns around the spiral, ever widening. In a longer work (that needs Deep Context) the Synopsis can become more heavy duty as well, extending into what you might classify as an executive summary.

If you have a software development background, as many participants here do, you might liken the structure of points 2 and 3 (the lede/synopsis) to a style of programming where you don’t cram everything into one line, but use top-down design. The idea is to keep the first sentence simple, not nesting details as subordinate clauses.

In languages like Haskell you can write it in the same order as we have here:

question = f x + g x + h
where
f = ⋯
g = ⋯
h = ⋯


Sometimes the order isn’t strictly to put the question body first with details following. Sometimes you might need:

let f = ⋯
question= f x


In any case, the actual question (stated as a question with a question mark!) should be easy to find, and you don’t have to wade through too much to get to it. In all cases, it should still be a synopsis paragraph that contains just enough detail to grasp the point of the actual question.

## Here is an example from the Sandbox:

The original version put a Context paragraph first, explaining the intended model of time travel. Then it had a breakhead for “The actual question” and if you’ve read these lessons in order you already know what I think of that!

Here is the first of two sentences from the “actual question”

If we assume that time travel is possible in our fictional universe and that the universe follows the above time travel paradigm, is it fair to assume that whenever the past is re-lived, it never occurs the same way twice - from minor behavioural/speech differences in people, to freak accidents killing people off unexpectedly, even different babies being born and thus the whole course of history changing (if you traveled back far enough).

Look how many sub-explainations are crammed into one sentence. To simplify, first I figured that the readership will already be familiar with the main time travel paradigms, so I only need to refer to it in a clear way. That means hunting around for a clear and succinct name for it, and then using this name and including a link so in the event that it’s not clear, someone can expand on it himself.

In keeping with the original structure, but simplifying the sentence itself, I broke the qualifying clause into a separate sentence.

Suppose we have a Branching Timeline paradigm, in which time travel into the past creates a new timeline. Is it fair to assume that any probabilistic events which might occur might be very different to what happened in the past in your original timeline?

This could be further hammered into the strict template form, but these are only guidelines and this is clear enough now. But for the exercise, let me try:

### Lede

Will a time traveler see random events play out the same way again?

This is incomplete, since we clearly identify necessary facts that are missing. But that’s OK, as the rest of the synopsis will provide these, hopefully in the same order as the urgency the reader experiences when wondering about them. I might also consider the clarification of the nature of randomness to be needed as part of the synopsis. In the original post, this was something the poster needed to be informed about in the answers, so it was under-explained in the question.

### Full Synopsis

Will a time traveler see random events play out the same way again? We are using the Branching Timeline paradigm, so the traveler arrives in a new timeline and will continue playing forward from that point. I’m referring to “true” random events, not changes due to the traveler’s own influence as a butterfly effect.

These suggestions were used in the Question and it earned a “Nice Question” badge.

# Lesson 5 — Abstraction (continued)

In Lesson 4, I said I saw two recent examples in the Sandbox.

# I've run into a few problems with this server

The question is a follow up inspired by an answer to another question. The draft post starts out by explaining the purpose of the servers, which has absolutely nothing to do with the real question. It continues with a recap of the previous post, which is that the servers are embodied in giant robots as a solution to another problem.

Then, and only then, does it come to the question, which is how to hide the giant robots.

The key here is to abstract the real question from the scenario under which it arose.

The fact that they are embodying servers is only tangentially important. What job the server is running is completely unimportant. So why is that given first? I think this illustrates the concept explained at the beginning of the previous Lesson: the question arises when working on a scenario, and the first reaction might be to brain-dump what you were working on and then append the question.

But the problem in the question is not necessarily bound to that specific scenario. Understanding the real issue will help you understand your question, as well as write a good post for Worldbuilding!

So, you have 60-foot robots that need to remain hidden from the general population. It’s important that the robots be active/“alive” because their underlying purpose requires them to have an energy source to run the servers, and engage in ongoing maintenance and repair of themselves so that they can continue filling their purpose.

The tech level is given: Just after WW-Ⅱ.

It’s important to give information that would exclude some ideas. Here, it’s noted that the robots don’t swim.

Now I think a good solution would be to put the robots on the ocean floor. Disallowing that seems like an artificial restriction just to make the plot work in the manner originally envisioned. So either the question was answered via comments while still in draft (so nevermind), or the question needs to elaborate more to explain why it’s limited to terrestrial environments. Are they of alien origin? Then why not set up in the asteroid belt or Mars or somesuch? If they’re not aliens, why are they so much more advanced then the people they’re hiding from?

In this case, I think the question needs more justification. If there was in fact a good explanation for all that, then a paragraph (as part of the Brief Context) can go into the story/background to that extent, as explained earlier in Lesson 2 (Why are they doing this?).

But at least we fixed the abstraction issue. Any readers want to give their ideas for a lede sentence, in the comments?