I perceive multiple distinguishable and yet interdependent components of any world. Things like:

  • individual creatures
  • societal constructs / interaction
  • geography
  • history
  • sources of strife
  • etc.

However, it is daunting to design all of them completely simultaneously. Is there a natural starting point?

And if so... Which of these is the most natural / easiest foundation to establish first when designing a world?

I guess I'm asking if one of these really flows into and facilitates the others.

  • 1
    $\begingroup$ possible duplicate of How can I break down the task of creating a world into manageable chunks? $\endgroup$
    – Styphon
    Commented Sep 19, 2014 at 19:58
  • 1
    $\begingroup$ @Styphon I strongly disagree. Our questions are quite distinct. Your answers do not suffice as answers to my question and my answers do not answer your question. Your question speaks of an organizational process of design. Most notably how to disassemble and later mesh all the puzzle pieces. My question is asking about what is the easiest, most basic building block. If my question is a duplicate of yours then surely yours should be closed as "too broad." Closing mine as a duplicate of yours would require closing ALL questions about non-specific design planning as duplicates of yours. $\endgroup$ Commented Sep 19, 2014 at 23:02
  • $\begingroup$ I'm not quite sure about this question, but "which is the most natural / easiest foundation to establish first?" honestly seems rather opinionated to me. $\endgroup$
    – user
    Commented Sep 19, 2014 at 23:26
  • 1
    $\begingroup$ @MichaelKjörling I see where you are coming from to an extent but I don't see this as a completely subjective or open-ended question and the answers have all pointed in the same direction. It is a narrow question that partly operates on the premise that all worlds can be objectively boiled down to the same core components that overlap in the same general manner. That objectivity about how the layers co-exist is the core question. Having any subjective element to the question would make World Building an extremely narrow site that wouldn't make it out of beta. $\endgroup$ Commented Sep 20, 2014 at 0:01
  • 1
    $\begingroup$ Maybe edit the question to ask whether there are dependencies, such that X always depends on Y (and thus designing Y before X is a good idea in order to have a consistent world)? That should, at the very least, do away with the opinion-ness, without invalidating answers already given. $\endgroup$
    – user
    Commented Sep 20, 2014 at 17:09

4 Answers 4


In the real world at least, geography (and climate etc.) is definitely the base.

From there you can easily lead on to what creature varieties and sources of strife. Sources of strife are always about resources until you get "intelligent" enough to have dogma; geography/climate determines what food sources, water, minerals, shelter etc. are available; and hence may be desired by others.

Social constructs are basically a combination of sources of strife and history. History is a the evolution of social constructs, external events, internal events and sources of strife. So history is pretty complicated and so are social constructs. In some ways things like geography/climate are just as complicated but the complicated details aren't really as apparent to a casual observer nor do they have as much impact on society.

So I'm basically suggesting you go from simplest to most complex in terms of the amount of details you'll have to provide to have it seem real. To detail that a bit more: start by working out a rough but serviceable geography/climate and then building up with creature/sources of strife (which often will be geographical features - for example adding in an oasis at this point of the desert). Now the reason you only did a rough physical world is that for the social constructs/history you may say oh it'd be nice if there was this canal here so that there could be trade between the (for example) Alemorids and the Zydia; if your geography is really hard built adding that is going to involve a lot more editing than if you have the broad brush only up to that point. Then once you've built everything, go back and fill in any holes.

  • $\begingroup$ Simple to complex makes sense as does how you explained the overlaps in your third paragraph. You talk about the less visible complexity of geography/climate. Are you also implying that it might make sense to design the geography fairly well at the start, build everything else on top, and touch up the complex aspects of geography during or after everything else?? $\endgroup$ Commented Sep 19, 2014 at 19:46
  • $\begingroup$ @Mark Yes. Edited with a bit more. Though actually a lot of the complexity of geography is just going to bore anyone except a [insert scientific speciality] reading it to tears; like we don't care how many types of dung beetles there are, most of us if we even think about it will assume there is something like that but unless it is important for the story that kind of thing just makes for ugly reading. $\endgroup$ Commented Sep 19, 2014 at 20:40

If you were to take a large enough sample of writers, you would probably discover that every possible starting point has been tried, and worked, for someone. Every storyteller is inspired in an different way from every other storyteller, and differently for each story. Every story is discovered via a new path that has never been traveled before, and will never be traveled again.

When starting a new story, game, etc., hopefully you have some starting point - an idea, a theme, a character, etc. Take your starting point, and ask yourself what that starting point requires. A character who is wizard is going to need a magical system in which he practices and a history - parents, birthplace, etc. A theme implies an appropriate conflict, which in turn requires a set of combatants and a context.

As you start to fill in details, it will be like tracing the components of a tree. If you start with a leaf, you will trace it to its branch, which will trace back to a larger branch, and so on until you eventually find all of the other branches, the trunk, and the roots. If you start at a root, it will lead you to a larger root, and then a large one, until eventually you find all the roots, the trunk, then the branches and the leaves.

My ideas are usually something along the lines of "What would happen if there was a person X living a world Y?" Where the properties of character X and world Y are defined only in the very broadest terms. From there, I can usually figure out how the story is going to end. Everything else must follow logically from those starting conditions to make the story interesting, compelling, and inevitable.


David Eddings claimed he always started with the map, but he wrote fantasy.

Sci-Fi I think would be different. I would think History and the current challenges would be more relevant. As in how did we arrive at where the story is taking place and why is it important to us.

Mysteries would center around the people(things) involved.

Is the story going have a large social commentary? Then societies and cultures would be the starting point. basically what kind of story are you trying to build, then identify what is important for that story and build out form there.

  • $\begingroup$ David Eddings didn't really write stories though. He wrote well but he basically had 5 characters and 1 story line that he constantly recycled. $\endgroup$
    – Tim B
    Commented Sep 19, 2014 at 19:31
  • $\begingroup$ I can't argue with that. He wrote one story and recycled it. But he did do fairly well, Janet Evanivitch is doing the same thing. :) $\endgroup$
    – bowlturner
    Commented Sep 19, 2014 at 19:36
  • $\begingroup$ Some social commentary but not completely permeating and in a fairly abstract, large concept way as opposed to specific current events. My focus right now is on fantasy novel writing. Sci-Fi may come into play on later projects. I like the way you split those up. It makes sense that Sci-Fi (as it is further from the way our society operates) depends heavily on how things got there and how the Sci-Fi aspects affect the tension in society. And I can see how in fantasy that would be lessened and shift focus to something else and the map seems logical. $\endgroup$ Commented Sep 19, 2014 at 19:38

From my experience in both three of fantasy, sci-fi and roleplaying games, the best start is always the history you want to tell (or play). From the history you get the events, the people, the terrain inside the zone you need, and you do not need to go further.

Think on Middle Earth. Do we know all of it? No, we only know the north-western part of it, because that is what Tolkien needed to tell his history. We known the languages on that zone, but do we know all the relations among them? No, because Tolkien only needed details for main characters' languages, and even there, mostly for naming details in Westron. Do we know about climates? really not, just some vague descriptions throughout the books, most of them not defining climate but the specific weather of the day the characters were somwhere.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .