Novels and RPGs have one thing in common – you have to describe a whole boatload of locations every time you play. As a result, every GM learns the basics of doing so very quickly. Unfortunately, once they achieve a level of minimal proficiency, most GMs never give this aspect of their craft a second thought. It seems relatively trivial in comparison with things like how to craft better plots, how to better capture and convey the nuances of personality, how to manage the real-world aspects of a game, and other headline-grabbing skills.
The majority of those who look past these priorities and actively try to improve in all areas of their craft quickly discover that the “rules” of a good descriptive technique for locations are largely unwritten and usually contradictory, and give up in disgust and frustration.
I don’t have any magic solution to take the chaos out of this situation. But, rather than give up, I say, let’s embrace the chaos and at least spell out the contradictions, and then try to assemble some guidelines to serve as the foundation of a solution.
Contradiction 1: Scope
There are three overlapping contradictions in the general advice available for the description of locations. Two of these proved too difficult to separate, dealing (respectively) with what descriptive elements should have Priority and with the sequence in which subsequent descriptive content should be presented. I was more successful at extracting the first from the general melee, the contradiction of Scope. How much detail should you include in your descriptions?
Priority 1: Economy and Concision
The less you have to say, the more easily digested and comprehended it is, and keeping your descriptions brief leaves more time for other things – like actual play. It seems fairly clear that less is more.
Priority 2: Comprehensiveness
I’ve written before about the declining imaginative capacities of modern times, and the solution of relying on visual aides to enhance the game as a way of compensating (‘The Gap In Reality: Immersion in an RPG Environment‘); The more you leave out of your descriptions, the more you rely on your player’s imaginations to fill the gaps – imaginations that may not be up to the job. So it is essential that you give those imaginations as much support as possible with fulsome and comprehensive descriptions.
What’s more, abridged and abbreviated descriptions can leave the game world a shifting chimera without substance and depth, a lightshow on the wall; more articulated descriptions shine light on the shadowy corners of the environment, giving the Gamemaster (or author, it’s all true for other kinds of writing as well) the opportunity to create depth and emotional context. The environment informs the personalities and influences of the occupants, so a comprehensive description achieves more bang for your descriptive buck.
Finally, there is nothing worse, as a player, than responding to a situation or scene only for that response to fail because of something being left out of the description. This frustrates players and puts an unnecessary strain on the game. There is, accordingly, a perpetual desire on the part of GMs to err on the side of comprehensiveness; if you describe everything (or at least mention its presence), there is no chance of leaving something out.
Guideline: Layered Poetry of Language
There are no simple solutions to this dichotomy of priorities. There are clear benefits to both approaches, both extremes. Striking a compromise can artfully harness the benefits of both, but risks introducing the inherent problems of both, as well.
How can you maximize the chances of getting the description right – just enough detail without saturating the players with irrelevancies? The best guideline I can offer is to learn to be poetic in your choice of language. By layering additional meanings into a few well-chosen words, you can suggest more than you explicitly show, jumpstarting the imagination without bogging down in specifics and tedious details.
Here’s one description of a city, which trends toward completeness rather than poetry:
It’s a very crowded city, with architecture crammed into every available space. Buildings crowd the streets as much as people do, jutting into the street in a haphazard manner. Most of the buildings in this part of town are three stories tall, constructed of red bricks or gray stones, and everything is unnaturally dirty and soot-stained. There is a dank and oppressive smell, and mould grows on many of the walls. The roofs are mostly constructed of reddish tile, though a few have slate tiles, and are angled steeply; you suspect most of having an extensive attic space for storage. From the higher stories of each building, gray stone balconies protrude, providing shade and shelter from the rain for those walking along the sides of the streets. There are no sidewalks. Clotheslines stretch above the streets from one building to another, connected to pulleys that enable clothing to be aired and dried after washing. Chalked religious icons are present on most of the doors, which are usually constructed of thick hardwood with heavy iron locks. The streets are narrow and roughly cobbled, not very level, and footing is uncertain. From place to place, deep pools of grimy and muddy water gather in depressions. The populace always seems to be busy, and generally look down as they rush from place to place, rather than meeting each other in the eye, almost as though they were all scared of each other. They typically dress in dark woolen cloaks and wear wide-brimmed leather hats of poor workmanship. Footwear is usually sandals, over thick woolen leggings or knee-high socks. Knife-sheaths and purses adorn leather belts to either side of the wearer. Carriages and wagons groan and clatter as they travel down the centre of the roads, swinging out to the sides of the streets only one must pass another; at such times, they pass mere inches from the faces of pedestrians who are forced to cling to the walls of the buildings to avoid being struck. Most are drawn by one or two horses, and evidence of their passage is perpetually underfoot. Every morning, barefooted convicted criminals wearing white smocks and red arrow-heads pointing downward scatter straw and hay on the streets to bind the dung, and each afternoon, they shovel the resulting mass into carts, adding to the crowding of the streets. Guarding the criminals are soldiers in bright blue uniforms with brass buttons which have been cast with insignia of rank, armed with swords, wooden truncheons, and whips.
But this description is only the beginning; there is no suggestion of commerce, or of the way people interact with each other – but they must be continually bumping into each other, if they are always looking down. So, to be comprehensive, you would need to add descriptions of street vendors, and perhaps a vendor haggling with a customer, and of a protocol that has people always pass to their left, and beggars with bowls begging for charity, and street urchins running from place to place in pursuit of a bouncing ball which they kick when they catch, and dogs without leashes and collars roaming the streets. And then you might realize that there’s no mention of windows in the buildings, or of the plan size of the typical structures, so you need to go back and insert that into an appropriate spot. And before you know it, your description is a page in length, and takes ten minutes to read to the players. But even without those details, you certainly get a vivid impression of the place.
This sort of description is the result of the GM (me, in this case) picturing the place, and describing what he sees in his mind’s eye, one thing at a time: Buildings, streets, traffic. It certainly conjures a vivid picture of the place, but Economy of language is clearly not a priority. In fact, it swarms with so many details that twenty minutes later, most will have been forgotten. And the chances of finding a picture that exactly matches this description is somewhere between slim and none, so the description either has to be compromised to match something that’s close but not quite right, or the GM has to rely on the description alone.
How would I go about rendering this image in a more concise manner?
1. Extract elements to convey with mini-encounters
Take out the wagons, and the passing foot traffic, and the street vendors, and the street urchins, and the dogs, and the prisoners & guards, and even the laundry drying and dripping from overhead. Each of these can be used as a mini-encounter in the streets, enabling the PCs to interact with the environment instead of simply looking at it. By breaking the information into smaller bites that can be delivered separately, you not only make them more easily assimilated by the players, you make them a little more memorable, and you make the players feel that their characters are a part of the landscape.
2. Extract anything that can go along for the ride.
A lot of the details – the balconies, the pulleys, the unevenness of the footing, the puddles, the dung, the straw, the way the people dress, even the smell – can be details attached to those mini-encounters. That means there is no need to detail them in the descriptive passage.
3. Extract any unnecessary details. It isn’t all that necessary to mention the attics, or the roofing materials, or the locks, or the religious signs chalked on the doors, as part of the initial description. These details can wait until one of the PCs wants to examine the objects possessing these qualities more closely. So I would separate them and put them in bullet points after that descriptive passage for easy location when I needed them – though some may creep back into the narrative in the next step.
4. Describe the impression, not the cause.
Finally, since you have such a vivid mental picture of the place, describe not the picture but the impression that the totality creates:
Buildings jut into the narrow streets, leaning and jostling each other for room like the people who jostle and bump into each other while rushing from place to place, their eyes perpetually downcast. Grime and mould decorate the walls of brick and stone beneath steeply-angled roofs. Straw, wet from the rain, makes a futile attempt to bind horse dung together for easy removal from the roughly-cobbled streets.
From there, I would launch straight into one of the “educational” mini-encounters: You dodge to one side as a ball bounces across the street, pursued by a pack of urchins, who ignore the occasional deep puddle of muddy water gathered into depressions in the road, splashing any who come too close. From behind you hear the clatter of an approaching horse-drawn wagon, and you can see another approaching in the opposite direction in front of you. Pedestrians hurriedly flatten themselves against the walls beneath the shelter of second-story balconies because the street is barely wide enough to permit the two to pass without colliding. [Pause for PC reaction]. …and so on.
By the time you have described two women yelling gossip to each other across the street from their third-floor balconies as they hang washing on lines strung between the buildings, and the haggling of the street vendor and customer, and the arrival of a prisoner detail with shovels, cart, and guards, and accosted the PCs with a beggar, and mentioned to each of the PCs that they have been jostled or bumped into, three or four times, you will have established all the details that were not extracted into bullet-point details, and the city will seem even more alive than that vivid picture assembled from the big block of text. Sure, it might now take two pages, and three times as long to convey that information, but the benefits of clarity and concision will be added to the benefits of being comprehensive. That’s the difference between a passable (comprehensive) description and one that uses the power of poetic expression to deliver a concise description with implied details.
Tip: I never write the artistic version “blind” – I always build up my mental impression using the comprehensive detail, taking notes as necessary. It’s much easier to move text around once it’s on the page than it is to maintain consistency and come up with the finished artistic description and mini-encounters straight off the bat; it’s like putting the cart before the horse and driving them in reverse. Get your details and consistent picture, but instead of writing it as a block of monolithic text, write it as notes that can be easily reformatted and reformulated into that finished product once your mental image is complete.
Contradictions 2 & 3: Priority & Sequence
That deals for the “easy” part of this article. The second part deals with the many-fold problems of which information to give, in what order, and in particular, what should come first and what should come last when you are describing a location. The examples used in the previous section should not be considered indicative of anything except one possible solution, described as a “Mood First” priority.
Priority 1: Mood First
This principle asserts that the mood and tone of a location provides a context, a filter, which – if presented first – can color every subsequent detail, sparing the GM the problem of integrating that mood and tone into those subsequent details. In other words, if you do this first, you have wider latitude with respect to the descriptive language employed subsequently.
Priority 2: Most Relevant Detail First
There is another line of arguement that states that the first thing to present in a location description is the most relevant single item. This can be especially true when it is something unusual but that the characters should be able to take in their stride, because it gives them time to consciously assimilate the information, processing this important factor while the GM is presenting the rest of the description. I have also seen this principle stated as “describe the elephant in the room”.
Personally, I have mixed feelings about this approach. While it ensures that the pertinent information is accessible to anyone who skims a narrative text – in other words, makes the description more useful to anyone other than the author, and even to the author when he has to search past descriptions for mentions of “the elephant”, a lot of the time it seems to get in the way. The players / readers focus on “the elephant” and its presence in the room to the point where they miss other pertinent facts. Players will often react by stating actions that are rendered foolish by additional description (“I draw my weapon and charge!” – “You run headlong into the lava river that I was about to describe before you interrupted me.”), leading to frustration and conflict between players and GM, or continually interrupt description that they don’t find immediately relevant to demand more details about “the Elephant”.
Frankly, I think that extracting appropriate keywords and placing them at the head of the description for the GM to use as ‘signposts’ when looking for the description weeks, months, or years later gives you most of the benefits of this approach and enables you to adopt the alternative approach:
Priority 3: Most Relevant Detail Last
Anything that the PCs should react to with an action should be the last item in your description. And there should always be something, no matter how trivial. Why? Because it provides a natural transition from narrative text to play.
Priority 4: Establish Iconic/Symbolic images at the beginning
Still another approach prioritizes what are sometimes described as “Iconic Structures” in the narrative. I have also seen this described as “Putting Onion Towers In Moscow”, and as “The Eiffel Tower says Paris”. The Statue Of Liberty is an unmistakable icon of New York City – that’s why it says everything that needs saying at the end of the original “Planet Of The Apes”.
This method takes advantage of associated memories and impressions to achieve concision, at least in terms of the wider picture – mention the icon on the horizon, and then move on immediately to the specifics of the local vicinity. And when you describe it that way, it becomes clear that this is a special form of the “Elephant In The Room” – the definition of “relevant” has changed, that’s all.
But that small change, because there are additional benefits and therefore additional justification for this approach, makes it harder to actually argue against this method. Where it fails is in a somewhat hidden assumption: that you want it to become clear, immediately, that the scene is Paris, or New York City, or Moscow, or wherever. I would agree – with reservations, that I will discuss in a moment – with the concept of using an icon to flag the specific point in the plot at which it is to become clear to the characters that this is where they are.
I mentioned reservations. There is another, hidden, assumption, and it’s a whopper: that the shared impressions and associated memories of those being triggered by this “shortcut” will be roughly parallel and consistent, both with each other, with those of the GM, and with those of the adventure / plot. What are the odds?
The greater the shared background of those involved, the greater – it can be argued – the chances that this assumption will be correct, because they will all have access to a common cultural exposure on which to base the shared impressions. The more differences there are in background, the greater the likelyhood that it will fail. Similarly, the greater the shared experience of the location, the more likely it is that the assumption will be correct – and the more divergent those experiences, the less likely that is. The more interests people have in parallel, the more likely it is to be correct; the more different those interests, the greater the likelyhood of a different impression of the subject. And finally, the more opinions people have in common, the more likely it is that they will share a perspective regarding that icon and what it represents; the more differences people have in opinion, the less likely they are to have a common view on that one specific subject.
I have one set of mental associations with Paris and France in general, a mélange of sources from things I have read, from watching the Tour De France, from segments on Top Gear, from Iron Chef, from Masterchef Australia’s French visits and French visitors, from various scenes in various movies and TV shows, from various documentaries, and so on. Two of my players don’t watch Masterchef, one doesn’t watch Top Gear (at least I don’t think he does), none of them watch the Tour De France – even though we have broadly similar tastes in movies and fiction, and overlapping tastes in non-fiction, for sure they won’t have exactly the same impressions of Paris and France. My mother recently visited France on holiday – and I’m quite sure that she will have a very different impression again. My aunt visited Paris many years ago (in the 1970s I think) and it’s dead certain that she will have still another impression. A friend of mine passed through Paris en route to a holiday focusing on the Castles of Germany and Austria; his very different interests to those of both my mother and aunt will almost certainly have produced still another set of associations. None of those are likely to match the typical view of Paris of a New Yorker, or a Mexican citizen, or a native of Quebec – to say nothing of the impressions of someone who lives in London, or even a French citizen! And they are all likely to be a different blend of accuracy and inaccuracy.
That’s an awful lot of risk. It’s so much risk that it undermines the very concept of this Prioritization approach. However, the change to using an icon to flag the specific point where it becomes clear to the characters that this is where they are acts to minimize this risk by restricting the value of the associations to secondary descriptive items after establishing the most significant descriptive elements.
1. You can’t assume that what you want to convey with your ‘icon’ is the association that it will create in the minds of the audience, except in the broadest possible terms.
2. Things left unsaid are gaps through which confusion, complication, and even possible discord, can flow.
3. The important details of your description are too important to risk in such a dangerous gamble.
And that leaves the Iconic priority dead in the water, so far as I’m concerned.
But that’s only half of this. What of the proposal that any Symbolic descriptive material should figure prominently?
Aside from having similar (though perhaps lessened) risks of different associations, I think that this approach is yet another example of “The Elephant In The Room” – again, simply redefining what “most relevant” happens to mean. And like that approach, players & readers are too likely to focus on trying to figure out what the symbolism means to the point of distraction. So, for slightly different reasons, I also reject this approach.
Usual Advice: From the General to the Specific
When you look up any “How To Write” books or pages on the internet, they will usually tell you to focus on the general picture first, and specifics second, especially if “Mood” is considered part of that “general picture”. Most of the time, this is good advice, but – in contradiction to most of what I have written so far in this article – there are times when there are distinct advantages to prioritizing specific information, even absolute necessities. If I want to describe a row of houses with a Martian War Machine rising above and behind them, the last thing I should be doing is spending any time describing the Houses and any lawns or gardens. The imperative of the need for an immediate reaction is so strong that it doesn’t matter WHAT the rest of the environment looks like.
So singular is this piece of description that even Mood is secondary; the presence of the War Machine will transform any mood that may have been present. The only reason to include a statement of mood at all is to contrast “before” with “after”. Either these things have been around for a while, doing what they do, in which case it will be expected that the houses are burned-out ruins, or this is a new twist in the plot, in which case it will be assumed that the scene is one of normal, everyday, life.
Another example of this phenomenon is that jaw-dropping moment when the Ship is first revealed over Los Angeles in Independence Day. There, a mood of normality is established, then skewed slightly, then reestablished through the kid “shooting aliens” with his toy gun – just to produce that jaw-dropping moment of discovery through the power of contrast.
Guideline: Hierarchy of Relevance
So, all components of a descriptive passage are not created equal. In every scene or setting description, there is a Hierarchy Of Relevance. Anything that comes after an item of descriptive of any given level in that hierarchy must stand higher in that hierarchy. If there is no direct-response trigger, the general rule (general to specific) may apply; anything likely to cause a reaction should normally be placed at the end of the descriptive passage, because the need to react will signal the end of the descriptive passage and a transition to interaction with the scene.
The more significant the direct-response trigger, the more description of location it makes irrelevant, to be included only for specific and intentional purpose.
That hierarchy is a key requirement to crafting a good location / scene description, making this a very important decision. It will tell you what should come first, what should come second, what should come last – and what should be left out, unless you can sneak it back in somehow.
Guideline: Layered Poetry of Language (again)
If information is to be left out of a descriptive passage because the elephant is not only in the room, it is charging you, then incorporating the most salient points of the lost description into the initial language becomes a secondary priority to be achieved through the artistic use of language. Instead of describing “A Martian War Machine rising up behind a row of houses”, you could speak of “A Martian War Machine blotting out the sun as it rises behind the row of single-story tenement housing and white picket fences”.
But there’s a limit to how much detail you can build into the description of the scene. Everything that I’ve included in that example is there for a purpose: a general description of the buildings and, yes, something that is symbolic of US Suburbia and small towns. They imply where this is happening, and permit the reader or players to make some reasonable assumptions about that location which will shape and influence their reaction to the primary descriptive item.
Technique: An Arc Of View, An Individual Perspective
Heading for the end-point of this article, still with nothing clearly established in terms of a “bible” for writing location and scene descriptions, only a couple of tentative and somewhat vague guidelines, and a technique or two.
Unfortunately, it was always going to be this way. There are too many situations, too many complex possibilities, to provide any hard-and-fast rules.
I have one more technique to offer. It’s not always the best choice, but it can work very well indeed at times, and should be in every GM’s toolkit.
The Fog Of Priority
It can be argued that what a character will notice first is anything that poses a direct and imminent threat; and that what they will notice first if there is no direct and imminent threat is something that is particularly relevant to themselves. A mage will notice the most obvious magical effect. A cleric or Priest will notice the most obvious religious-oriented detail. A fighter will look at the dominant tactical impact of the terrain, and so on.
Rather than providing the description as though it were one collective perception that they all experience, try this: General introduction & Mood, Character #1 insight, Character #2 insight, (and so on), until each character has noticed the thing that is most relevant or important to them. If there’s more to be said, you can do a second round, though in general, one round (with a couple of passages ‘snuck in’ using the techniques demonstrated earlier) is probably enough to cover everything that matters.
This approach has the effect of obscuring details that the characters are not paying any attention to. I call this effect “The Fog Of Priority”, and so long as the GM is prepared to provide additional description of whatever the player says he is looking at more closely, and ensures that all the critical information is delivered in his opening salvo, this technique integrates the influences of the characters into the description.
Guideline: Expansiveness, Mood, Essentials, Triggers, and Clarity
My final stop on this tromp through the narrative wilds is to offer one final guideline for writers and GMs to follow: before you start to write a description, set the goals that you want that description to achieve.
- Expansiveness: Long, flowery descriptions have their place. Short, compact descriptions have their place. What is the right length for this occasion?
- Mood: What is the mood, the atmosphere, the overtone, that you want the description to convey?
- Essentials: What is the absolutely essential information that the description has to convey?
- Triggers: Reactions form a natural transition from description to action or to roleplay/conversation, so you should always try to end your description with something that will trigger a reaction – but what reaction do you want to trigger?
- And, finally, when you’ve written the description, Clarity: How clear is the meaning of what you’ve written? Do you start talking about one thing and interrupt it to describe another? Is there any possibility of the description of one element being misapplied to another?
Hey, what do you know? We’ve arrived at a hard-and-fast rule for most occasions, after all :)