I read a neat ttrpg article by JD Clément called Against Abstraction, Or, Tags are a Mind Virus. You should go read it. It’s very well-written and thought-provoking even if I disagree strongly with it.

Clément is arguing for a naturalistic, diegetic way of describing things in ttrpgs. Instead of applying mechanical “tags” like unbreakable, one-handed, or incorporeal to objects or characters within the game-universe, Clément argues it is better to simply describe them as they are:

“This is, at its core, a call for better prose and fewer shortcuts.

These types of tags are the least offensive and the most easily remedied. Just remove the brackets, get rid of the italics, and use the descriptor tag as the adjective it was always meant to be. A sword (one-handed, holy) becomes a one-handed holy sword, or a broadsword blessed by the saints. These are in-universe facts about the sword, facts with weight and meaning not just to us as players, but to our characters. Give me the thing itself.”

First off, I have to say that Clément is absolutely correct that this would result in more appealing, immersive, “better” prose. As a writer, reader, and player of TTRPGs, it’s undeniably cumbersome to read about what should be unique and fascinating magical artifacts, spells, powers, etc. but in reality feels more like a spreadsheet.

No argument from me about the prose.

Here’s the thing though–I don’t read TTRPGs to experience good prose. That’s what fiction is for.

Terms of Art

First, what Clément is criticizing as “tags” would be better understood as “terms of art.” A term of art is a word or phrase that has a very specific meaning within the context of (in this case) a TTRPG, distinct from any meanings it may carry outside of that.

The word “cantrip” has historical meanings associated with trickery, witches, and chanting. In Dungeons & Dragons, however, it’s a term of art referring to a specific kind of spell with its own mechanics.

Clément writes of game mechanics related to firearms, for example, that “We can assume our reader understands the basic facts about firearms.” While I actually think this is debatable, even if we accept it as true, is that true of cantrips? Do most people know what that means outside of D&D? Does it even mean the same thing outside of D&D? (Obviously, the answer is no.)

The response would be that you still don’t need a “tag” or a term of art for this. The common characteristics of these spells could simply be described diegetically (in-universe) with something like, “A simple spell, one that any mage can cast at-will, without needing to prepare it first.”

A few problems immediately present themselves. One, I’ve also had to eliminate the term of art “Spell Slot” since that’s another abstraction. Meaning that we’re assuming the reader understands the concept of “preparing spells” as it pertains to D&D and other games like it. If they don’t, then that description really doesn’t help, does it?

But the problem is much bigger than that. A ttrpg rulebook, in theory, exists to 1.) teach people the rules of the game, and 2.) be a reference for players as they’re actually playing.

This is always a balancing act. A rule that’s written in such a way as to be easy for new players to grasp can sometimes be more challenging for veteran players just trying to get a quick reference, and vise-versa. (The Alexandrian explains in more detail here.)

Basically, a term of art exists to make rules easier to reference. If you have some degree of system mastery (or even just “system experience,”) you see the term, you know what it means in terms of mechanics, and you move on.

These types of diegetic, prose-heavy descriptions annihilate that utility. Even worse, they don’t even have the tradeoff of being easier for new players to understand, since they don’t actually teach you how to play the game.

Rules Are Important

Clément doesn’t exactly argue that rules aren’t important/don’t matter, but does brush them aside throughout the article.

Taking their “broadsword blessed by the saints” example, what does that actually mean in game terms? What makes it different from a broadsword not blessed by the saints?

This is where rules like “+2 damage vs undead” come in. Clément finds rules like this “kind of boring” and says that they “wouldn’t use them,” but doesn’t entirely dismiss them. The problem is that they don’t bother to explain how a rule or mechanic is meant to emerge from, fit in with, or interact with their idea of “diegesis.”

According to them, you can still have rules like “+2 damage vs undead;” it’s just the tags or terms of art that are the problem, that get in the way of diegesis, of the fiction.

But without the term of art, how are you supposed to communicate the rule to the player?

Rules like “+2 damage vs undead” exist to provide consistency to things like holy weapons (or any number of in-universe abilities, traits, objects, powers, or anything else that requires game mechanics). Without them, relying solely on diegetic descriptions, a gaming group must determine either through collective agreement or GM fiat what that description actually means in game-terms every single time it comes up.

Clément seems to find this appealing. (We’ll get into why I think that is a bit later.) They write:

“If we cannot agree on the diegetic facts, we discuss until we reach a consensus, often by exploring the fictional world in character and discovering what is and isn’t true, or by asking questions and creating answers. We are playing to find out in the most literal sense. You do not need tags to do this.”

This sounds nice in theory, but it’s rendered frustrating by the article’s failure to explore how you could have rules without tags. They state that you can still have rules like that, if you want to, but then have nothing to say about how you are supposed to use their prose-first, diegetic descriptions to communicate them. It’s an enormous gaping hole in the article that becomes apparent the more you try to put their advice into practice.

Taking the article utterly at face-value, it seems to suggest replacing tags with diegetic descriptions and then relying on those descriptions to point players to the rules, which are presumably expected to be found…somewhere. Since Clément isn’t advocating getting rid of rules entirely, this is the most straightforward reading of what they’re suggesting we do about all this.

But think about that. What effects would that have on an average gaming group?

You’d experience whatever pleasure might come from reading a tiny snippet of literary writing. Then you’d have to interpret it mechanically and look up the rule anyways, except without the benefit of a term of art you can quickly reference.

A “sword blessed by the saints” is potentially interesting. But do all holy weapons have to be blessed by the saints? Does it matter which saints? Are there other ways weapons can become holy? Are we certain that every single diegetic, in-universe description of a holy weapon the writer or designer comes up with makes it 100% clear to every single player that they should refer to the rules about holy weapons (wherever they may be)?

Terms of art and tags exist to make rules easier to use. They give utility to your rulebook. While “blessed by the saints” can be open to interpretation and debate, terms of art cannot. (Not if they’re well-designed and explained.) You find a weapon, see the “holy” tag, and immediately either know what it means in mechanical terms (if you have some system mastery) or know exactly how and where to look it up.

Tags make sense when they refer to a consistent rule that will be used in multiple places. Clément’s article obfuscates this by giving nonsensical examples of tags. Let’s examine a couple:

• “The blade’s golden filigree catches the light, untarnished after all these centuries.” (Ornate)
  • Why would you need a tag for “ornate”? What rules are actually required for “ornate”? The only ones I can think of would be rules related to buying/selling/monetary value. Either the item could simply have a high monetary value, or you have some kind of specific mechanic regarding how to sell ornate items, who they can be sold to, etc. In a game focused on mercantilism, that’s probably not just useful but necessary. In a normal RPG where mercantilism is ancillary to the primary game loop, why would you ever need this tag?
• “The drone, assembled by master dwarfs in the sacred forges of Zerek-Uth, struggles to differentiate friend from foe through its cracked ruby optics.”(Glitchy)
  • Again, why would “glitchy” have a tag? Glitches are so abstract already, and can stem from so many different sources, that having a single “glitchy” tag actually raises questions, which is the opposite of what tags are supposed to do. Do all glitches stem from cracked ruby optics? Obviously not. What rule is “glitchy” pointing to?

On the other hand, there are examples where a tag pointing to a specific rule clearly would be helpful.

• “Your warrior can hold the sword and find it most comfortable to use with a single handed grip.” (One-handed)
  • Cool. Does this mean that it can only be wielded one-handed? A tag like “one-handed” usually exists to tell the player that they can use their other hand for something else, either at all, or without a penalty (and the corollary, that some weapons forbid using something in the other hand). And what does “most comfortable” mean? Is there a penalty for using it in a less comfortable way? Or is this all fluff and none of it matters anyways? (Which might be fine, but in a lot of games, this stuff does matter.)
• “The breastplate, blackened by wyrmfire, draws out the ferocity of its wearer, turning their mind to violence and bloodshed.”  (Cursed)
  • Curses can have all kinds of different effects, but having “curse” as a type makes it so that players have consistent and clear means of dealing with them. (Remove curse scrolls, for example.) It can also be important for magic systems/in-universe lore–perhaps all curses stem from the god of dark magic? Or they result from an unnatural manipulation of whatever magical energies make magic possible?
• “…only the cleric’s silver mace seems to harm the banshee, as all other attacks pass like wind through fog.” (Intangible)
  • This is the most interesting (and telling) one, in my opinion. What rule would you intuit from this? Banshees can only be harmed by silver, or by clerics? Or by silver maces wielded by clerics? The rule that the “intangible” tag points to tells you, very clearly, what items can affect intangible creatures.

The reason the silver mace is so interesting is because this isn’t a rule that should be handled by a tag; this is a description. This is something you’d say during a session when players encounter a banshee for the first time–it’s the descriptive text emerging from the intangible tag that points to the rules about intangible creatures.

Clément’s whole argument in regards to tags and prose is a solution in search of a problem. The issue of “better, more immersive prose” is a solved problem. It’s called flavor text — the often-italicized, diegetic, literary and descriptive prose found at the bottom of Magic cards and at the beginning or end of all kinds of RPG stat blocks, from monsters to items, since basically the beginning of the hobby.

So the stat block for the sword would be something like:

Holy Sword (one-handed, unbreakable, holy)

A broadsword blessed by the saints.

Say you then have another holy weapon–a unique one. We’ll call it Demonslayer (original, I know). It could then have a stat block like:

Demonslayer (two-handed, holy)

A gleaming white claymore once wielded by Duncan the Demonslayer in the Demon Wars.

Now, imagine we didn’t have tags and simply used the diegetic flavor text. How do we communicate to the player that Demonslayer uses the same mechanics as “Holy Sword”?

The rejoinder might be that it shouldn’t use the same mechanics, but that’s a matter of design goals and preferences. Suppose for the thought experiment that for whatever reason we do want these two weapons to use the same mechanics. What better way is there to communicate that than with a simple, one-word tag?

You could simply put the rule into each stat block, and in fact I encourage that. However, having the tag still helps because it instills system mastery. After seeing it a few times in a few different stat blocks, players and GMs will learn the mechanic for “holy.” Then, in the future, if they ever need to use that mechanic, they will immediately know what it means.

There is no reason not to use the tag if the tag is going to refer to a consistent rule.

I hope the problem is obvious. By removing tags, you jettison utility and gain nothing.

So, I must ask…why? Why are we even having this discussion? Why is this non-solution to a non-issue a thing anyone cares about?

Abstraction vs Dissociation

The choice between diegesis and mechanical abstraction is false. Clément’s article conflates evocative prose (the kind that is ideally found within flavor text) with diegesis. This false conflation relies, in my opinion, on a failure to comprehend the difference between associated and dissociated mechanics.

It always feels a bit strange to see RPG theories that were basically answered by The Alexandrian decades ago, but it keeps happening. This time we’re talking about dissociated mechanics and abstraction; you should go back and read the OG article by Justin Alexander if you’re not familiar with the terms.

For those who don’t want to, in brief:

An associated mechanic is a game mechanic that has a direct correlation to something in the game’s fictional universe, to something diegetic. Example: a sword receives +2 damage vs undead because it’s been blessed by the saints. You, the player, understand “+2 damage vs undead,” while your character understands “blessed by the saints,” but fundamentally these two understandings correlate with each other.

A dissociated mechanic is one that has no diegetic connection to the game world at all. Metacurrencies like luck/fate/inpsiration points make an easy example; when you use Inspiration to negate a mechanical check (or choose the outcome of an action, depending on the game), this has no association with anything diegetic. You the player understand that you used a meta-currency to change the reality of the game world, while your character understands…nothing. They have no knowledge at all that anything happened that wouldn’t have happened otherwise.

Dissociated mechanics aren’t inherently bad. Most roleplaying games have at least a few of them; things like experience points can become dissociated if you use XP you gained from killing goblins to raise your Speech skill, for example.

But a preponderance of dissociated mechanics is one of the primary signifiers of a storytelling game, rather than a roleplaying game. In a storytelling game, players have a degree of narrative control not available to those within traditional (and even most non-traditional) roleplaying games. While characters in RPGs might get occasional Fate or Inspiration points, storytelling games are fundamentally built upon such mechanics.

In fact, if you were trying to describe what a storytelling game looks like, something like:

“…we discuss until we reach a consensus, often by exploring the fictional world in character and discovering what is and isn’t true, or by asking questions and creating answers. We are playing to find out in the most literal sense.”

…wouldn’t be a bad start.

Wait.

Oh God damn it, not again! You’ve gotten your storytelling game in my roleplaying game!

Storytelling vs. Roleplaying, Again

I’ve talked about this before. There seems to be a contingent of tabletop gamers who prefer games where players have more narrative control than traditional roleplaying games, and to one degree or another argue that narrative control mechanics and other dissociated mechanics in general are desirable in roleplaying games.

That’s fine if that’s what you like, but for many people (myself included), dissociated mechanics ruin the act of roleplaying. In the context of an RPG, roleplaying is about using game mechanics to make choices as my character. I can’t do that when the mechanics don’t correspond to anything in the real world.

So when Clément writes:

“Ambiguity and negotiation at the table is fruitful. We are constantly re-negotiating the reality of the game world already (that’s what discussion and narrative is, after all). By relying on descriptions rather than prescriptive tags, we encourage a more natural collective narrative, rather than one that relies on the external mechanics of a drone being glitchy to define what a glitchy drone can and can’t do.”

–my hackles are raised. I don’t want ambiguity and negotiation at the table. I want the GM to tell me what’s going on in the world, and I want to roleplay how my character acts and reacts. I need a clear understanding of the game world to do that.

Roleplaying does not inherently involve “re-negotiating the reality of the game world;” it can, but ideally that is only when something has broken down. It happens when someone has a faulty understanding of the game world and/or the characters in it which needs to be clarified or corrected. (Example: a player narrates their character taking a nonsensical action like leaping off a cliff. Either they or the GM has miscommunicated or misunderstood something, such as the height of the cliff or an intention to use a magical flight-bestowing item, necessitating clarification and renegotiation.)

Constant dialogue and negotiation to create a natural collective narrative is what a storytelling game is. In a roleplaying game, I can’t roleplay my character the way I want to if I have the responsibility (or even ability) to re-negotiate the game world itself. It leads to strange storytelling and narrative skew, and an example late in Clément’s article demonstrates how.

Towards the end of their article, Clément has a very useful discussion about hyperdiegetic rules vs hypodiegetic rules. Hyperdiegetic rules are rules of the world, while hypodiegetic rules are rules in the world. The example given is a hyperdiegetic rule that emerges through play–a player wishes through magical means that dwarves in that world cannot use shields. This becomes an incontrovertible fact of the universe–dwarves cannot use shields. It’s as if it was written in the rulebook.

Now, in a traditional roleplaying game where the GM controls the world and the players control their characters, it would be up to the GM to associate this mechanic with the game world. Perhaps dwarves are magically compelled to drop any shield they touch. Perhaps shields shatter in their hands. Perhaps shields drain their life-force.

Each of these possibilities raises further questions–can dwarves of significant willpower resist the magical compulsion? That would override the hyperdiegetic rule, so probably not, but it would feel strange to players. Does a plank of wood become a “shield” if a dwarf uses it to block a blow, or does the item have to have been constructed as a shield by its creator? Etc, etc, and the GM will have to make rulings on these and more questions. That’s an inherent difficulty with trying to associate dissociated mechanics.

Leaving that aside, the actual problem comes about when Clément tries to show us an example of this “negotiation” of the game world in play. Instead of the GM associating the mechanic, it is left up to the player at the table who happens to be playing a dwarf, Tharg.

“Tharg’s player chooses to roleplay this as a deep cultural disgust with shields,” Clément tells us. While this is a rule “within the world,” Clément tells us that it cannot be a hypodiegetic rule, because we already know that “dwarves cannot use shields” is a hyperdiegetic rule, based on the Wish spell.

“It’s not a law, it’s a fact – dwarfs can’t hold shields anymore. If it were a hypodiegetic law, Tharg would merely need to pick up a shield and reveal that this law, like every other, is held together by ideology. But he can’t. If I were to say he does it anyway, that would break our diegesis – I wouldn’t be playing by the rules we all agreed to play by. It’d be kind of a dick move.”

While it’s true that Tharg’s player choosing to pick up a shield anyways would be breaking the RPG Social Contract and “kind of a dick move,” the bigger problem is that the choice to leave the resolution of this rule up to “negotiation” has created a dissociated mechanic.

The first and obvious objection is that there is absolutely no cultural disgust, prohibition, or taboo so powerful that there isn’t someone out there who will break it. In any realistic world where dwarves behave as we understand sentient people to behave, there will still be dwarves using shields.

But more fundamental is that the process of choosing to resolve this mechanic in this way has created a fundamental disconnect between how the players understand the world and how their characters understand it. By the rules of the game, dwarves cannot use shields; however, in the game world, they absolutely can, but choose not to.

Roleplaying a deep cultural disgust with shields sounds cool. As does roleplaying a character who flaunts taboos, or breaks them under extreme duress and then has to wrestle with that. None of that can occur in this game, however, because the cultural taboo is fundamentally dissociated from the game mechanics. Roleplaying has actually been stifled, and immersion in the game world is broken.

This is where a clear rule, tag or no tag, would come in handy. If an action becomes impossible through magical means, then an irresistible magical compulsion stymies those who attempt it.

You could argue that problems like this could be avoided by more careful negotiation at the table, but it isn’t always that easy to spot dissociation until rules and diegesis start interacting in complex ways. And, many players and GMs are not even aware of dissociated mechanics, or are not as bothered by them as some of us are. Thus, there is always a risk for these types of incomprehensible situations to occur that will actively harm certain types of RPG players.

And that’s why it’s important to distinguish storytelling games from roleplaying games. In a storytelling game, problems like that of dwarves and shields don’t really matter–the hyperdiegetic rule (dwarves can’t use shields) acts as a story prompt that spurred the diegetic explanation (dwarves have a deep cultural disgust with shields). In a storytelling game where the primary purpose is to create a narrative, that’s generally fine.

But in a roleplaying game, associated mechanics define the act of roleplaying. You cannot meaningfully roleplay with dissociated mechanics, because your decision as player will be inherently disconnected from your character’s decision in-universe.

It’s not just that I’m a different type of RPG player from Clément–it’s that Clément’s suggestions actually directly hamper their stated goal of achieving greater diegesis.

The problem is a flawed, incomplete understanding of diegesis. It’s not enough to have “good prose.” For true diegesis, you need game rules that are associated with the game world. You need a world that is believable and consistent.

Tags have absolutely nothing to do with whether that is true or not.

For Abstraction

I want to end by talking about why I like abstraction, and why designers might intentionally use it in their games.

First: abstraction is a separate issue from association/dissociation. A mechanic can be abstract and associated, or abstract and dissociated. (For example, XP is always abstract because it isn’t something that the characters know about diegetically, but it is associated with the character’s experiences and achievements, to greater or lesser extents.)

Clément writes that tags and other abstract game mechanics “force themselves onto the diegesis rather than arising from within it, a way for designers to narrow the field of possibility by defining what specific adjectives mean within a hypothetical game world before it ever reaches the table.” We’ve already seen that this statement is misleading–the field of possibility is only “narrowed” in the sense that the game world is constructed before the PCs step foot in it, and the PCs do not get to “negotiate” that game world during play. For a lot of RPG players, that is a good thing.

Let me be clear: there is a spectrum between STGs and RPGs. Games often blend them (for example, I could see interesting roleplay coming out of narrative-control and game-world-negotiation gameplay if you were explicitly roleplaying some kind of god or deity). The problem is when designers refuse to acknowledge this spectrum, and the reality that there are people who sit comfortably at one end of it.

Clément’s ideas are interesting and may be fruitful for a specific kind of game. I’m not sure if I personally would enjoy that game, but not everything needs to appeal to me.

I consciously, purposefully enjoy games where I have little-to-no control over the game world. I do not want to re-negotiate the world. I want to roleplay my character and have the world react to that. On the other side of the GM screen, I want to create and play the world, and figure out reactions to the interesting things my players do. It’s what I like about GMing. I do not want them stepping into my realm.

It’s fine if you wouldn’t enjoy playing in that kind of game, but vast numbers of RPG players do prefer games that way, whether they can articulate it or not. It’s essentially the default, original form of roleplaying.

The medium can change, adapt to and incorporate new ideas, but that doesn’t mean those ideas are better. They’re just different.

Abstraction is often an unalloyed good in RPGs. The word “near” can have a lot of meanings depending on what you’re referring to. Elmhurst, Illinois is “near” Chicago, but not in the same sense that my bottle of Coke Zero is “near” my laptop right now.

In Shadowdark, however, “Near” is a term of art referring to a category within the game’s abstract distance/positioning system. It’s not tied to any specific number of inches or feet or meters, but within that game and that game only, a creature being at “Near” range, or a weapon having the “Near” tag, produces meaning through game mechanics.

Is this an abstraction? Absolutely. There’s no way to know through the mechanics exactly the distance between two creatures at “near” from each other, nor would that distance be consistent across multiple instances. What actually matters is that they can hit each other with weapons that have the “Near” tag.

The prose version of this would be something like, “Close enough to hit with a melee weapon.” That might be better prose, but again–why are we holding up better prose as some kind of priority in an RPG rulebook? We’re not writing a novel, here!

Abstraction allows us to say “these three creatures are all within near distance from each other” and move onto the interesting part–the battle.

Abstraction allows us to create situations and gameplay that would be torturous without it. It wouldn’t be fun to describe every single chalice and gem and coin in a treasure chest, even if it would be more interesting to read in a novel–216gp worth of treasure lets us note it down and then get back to the interesting gameplay of what we actually do with that treasure. In a hexcrawl, we don’t need to describe every thicket or grove–knowing that we are in “forested” terrain is enough to hang some gameplay around hunting, setting up traps, tracking, and navigating as we travel through the wilderness.

Tags, as stated, clarify rules, reward system mastery, and create a consistent game world. They enhance diegesis when used right. And they do not conflict with the desire for more immersive, diegetic, descriptive prose. You can have both, every time, all the time.