Characters are also an expression of setup/payoff, though I’ll admit that I didn’t realize this at first.

Basically, ‘agency’ means that the character is capable of making choices within a story. This is to differentiate them from other elements which are mostly static unless acted upon (by characters).
Where does the setup/payoff come in? Well, a character is the setup and their agency is the payoff to that agency. Though this is the super abstract layer and mostly irrelevant to storytelling unless you want to be super pedantic.

A character’s decisions are governed by who they are, what they can do and what they want. These can broadly be described as ‘personality’, ‘abilities’ and ‘motivation’. Though there are multiple ways to approach these (like with Story Structure). I’m gonna highlight the one I think is best.

I’m working off of the Chronicles of Darkness to determine this, though I’ve already diverged in a few key ways. The ideal ‘character structure’ looks like this:

• Core (or Anchor/Virtue)
• Crack (or Vice)
• Traits
• Perspectives
• Relationships
• Attributes
• Motivation(s)

The ‘Core’, or ‘Anchor’^A) or ‘Virtue’ (if you prefer to stick to CoD terminology)^B) I define as the ‘default state’. Basically: no matter what choice is presented to a character, they’re likely to go ahead with one based on their core.
The ‘Crack’ or ‘Vice’ is a critical character flaw which is like the easy way out. If presented with choices where this crack comes into play, the character is strongly motivated to choose the one that aligns with their crack.
‘Traits’ is where all the flavoring for a character goes and it can be as detailed or simple (or non-existent) as desired. However, the traits should ideally flow from the Core/Crack. They can conflict but they don’t have to. I’d also place things like occupation and hobbies here.
‘Perspectives’ is whatever the character thinks about various things. I’d consider this one optional, though it can help in further defining a character and figuring out how they respond to certain things. This includes self-image, politics, philosophy and so on.
The ‘Relationships’ is also optional and simply defines their relationships with other characters. It’s similar to Perspectives, but more focused on other characters.
‘Attributes’ is pretty straightforward as its whatever the character is capable of doing. Knowing how to hack a computer isn’t a personality trait – it’s a skill – so it belongs here.
‘Motivations’ is based around the character’s goals. These are the most fluid part of the character, as they can change more often than any other part of them.

The Core and Crack should be defined using a single word (such as ‘Duty’ and ‘Pride’) though I don’t think there’s an issue with using multiple (‘Duty’/‘Honesty’). The only risk is it becoming incoherent. It also really doesn’t matter what it is (though adjectives are better) as long as you know what it means and can explain it.
Another thing to keep in mind is this: while I’m using the words ‘Virtue’ and ‘Vice’ that doesn’t necessarily mean the Core has to be positive or always manifest positively (and vice-versa for vice, pun unintended).
The virtue ‘duty’ could result in a character betraying a friend, for example. The vice ‘greed’ might result in them demanding to be fairly compensated.
The Anchor and Vice don’t even necessarily have to conflict. They could be in concord. Or reversed. Maybe a character’s Anchor is something like ‘Sadistic’ while their vice is ‘charitable’.

I’ll create three characters to demonstrate. Note that prior to me figuring out this system, it was much harder to figure out how to make characters.
These ones only have the Anchor, Vice and a few traits defined, to show just how minimal the system can get.

All of these should already give a pretty good picture of how they may act in a certain situation.
For example, let’s take Irina. Let’s say that she was given an order. Due to her duty, she’s likely to accept that order. But what if it’s to clean the toilet? Well, in that case she’s likely to object due to her Pride. She’s also likely to take comparisons between herself and others very seriously, either boosting or damaging her ego in the process. However, if she’s told to do something illegal and degrading for the sake of boosting her career…she’s likely to refuse and report it, because her duty makes her unwilling to do something illegal and her pride makes her unwilling to do anything degrading.
Jonathan has the same Core and Crack as Irina, but it is likely to manifest very differently in his position as a lawyer. For example, rather than being willing to follow orders (the law), he might feel that it’s his duty to do whatever he can for his client, but his pride makes him unwilling to admit that. His pride also makes him unwilling to lose any cases he takes.
Shirley is a radically different character and you’ll notice that in her capacity as a waitress, her crack will rarely if ever turn up. But if her restaurant were to be attacked by an armed robber, she’d be one of the first to run away…or she may feel compelled to listen to the armed robber to avoid getting hurt. Her honesty could also manifest as her being insensitive to customers.

I think the flexibility of this system speaks for itself. Additional traits (like chatty, air-headed or aggressive) help to flavor the character and to differentiate them further from each-other. Maybe Irina is also air-headed and aggressive, compared to Jonathan who is polite and patient.

One of the other reasons I favor this system is that it lends itself well to automation and character standardization. Basically…you can use it to create characters that you’ll understand well automatically and you can use it to describe all characters in a story.
It may even be useful for story analysis, but that would be me getting ahead of myself.

Among Western writers, there is a high emphasis on motivation for determining characterization. While I think it’s important…I get the impression people end up missing what actually makes characters who they are.

The main problem is that a motivation is simply a goal that a character has. It tells us nothing about who they are. At most it tells us about what they believe in (which should ideally stem from who they are and not what they want). These motivations are also often highly specific to a given story and – as I theorize – are the main reason some characters start to feel ‘off’ in continuations.^C)

A previous version of this article detailed an idea I had (before I really figured out how characters worked) where you can determine the personality of a character based on their memories. Those memories would then shape who they are as a person.
A ‘memory’ is essentially a mini-story which explains something about the character. For example, Mary used to dress up as a princess for Halloween which ended up shaping her desire to be pretty and her ambition to be a model.

The issue I encountered was that this still doesn’t tell us who they are or why they make the decisions that they do. Though they can be useful for other means. I think they’d be best used for crafting the backstory of a character (which we build backwards from their characterization) or for coming up with a few extra ‘fun facts’.

This section is for additional character-related analysis that I’m putting together mostly on the fly.

I’ve heard characters be placed into these two broad categories before and, honestly, it’s completely worthless.
Well, maybe not completely but definitely mostly worthless. Similar to Wattsonian VS Doylist analysis.

The idea here is that ‘reactive’ characters are those who mostly react to the happenings around them and then act accordingly. ‘Proactive’ characters are those who are more proactive. They don’t react to the happenings around them as much as actively make them happen. When I heard this, it was stated with confidence that proactive characters are universally better.^D;E)

The issue is that it’s extremely difficult to determine whether a character is reactive or proactive, because you can interpret a lot of reactions to things as proactive actions. Alternatively, proactive characters could just be reacting to something else.
That said, this isn’t completely worthless as a writer might find it useful to see whether their characters mostly react to events or not. But I don’t think it’s good or bad either way, though a ‘reactive’ character could indicate a weak one.

This is one of my ‘crack’ theories and its usefulness is To Be Determined, as I’m in the process of trying to figure out some more character-related stuff.

A stable character is one who can be ripped from their context and placed into a new one without feeling out-of-place. An unstable character cannot.

This might be useful to determine whether a character actually has proper characterization: basically, if you can reliably predict how a character is going to act in any situation presented to them. If a character is not properly characterized, they’ll be dependent on the surrounding context to sustain them. This is also why, I think, some characters just feel ‘off’ in sequels, cross-overs or other such stuff, while other characters still feel like themselves no matter where you put them.

Another ‘crack’ categorization. I won’t elaborate on it too much, since I came up with it recently.

Tethered characters are those who are deeply tied to the story they belong to. They aren’t necessarily unstable, but ripping them from their context is undesirable anyhow. Untethered characters are those who can be placed into virtually any context.

To give examples: Indiana Jones can probably be placed into any context and he’ll mostly be fine. Meanwhile, the antagonist and protagonist of Daybreakers cannot be placed into any context, because their characterization is directly tied to the plot and theme of that film.
This may or may not be a useful means of classification and, in fact, I’m not really that sure about whether a ‘tethered stable character’ could exist.

I’ve come to the conclusion that this is a mostly pointless way of categorizing characters. The reason being that an untethered character doesn’t technically exist: every character is tethered to a differing degree.
I’ll use the example of Indiana Jones. Indy is a mostly untethered character: the setting, location and what’s up can vary wildly, but he’ll continue being Indy. However…Indy cannot be ‘native’ to the stone age…because professors and whips (presumably) did not exist back then. Neither did anything else surrounding his character. So, if you want an Indiana Jones who exists in the stone age…you will have to adapt him until he fits into that context.
This shouldn’t be confused for plopping a character into a setting. Basically…if Indiana Jones suddenly appeared in the stone age, you could still theoretically predict his actions…even if he isn’t ‘native’.