Ask a room of writers what an archetype is and you will get two answers that sound alike and mean opposite things. One camp treats archetypes as ancient furniture of the mind, inherited shapes that every culture is obliged to produce. The other treats them as a filing system, a set of labels that let a dozen people building the same story agree on what a character is for. The gap has been argued over for a century, and the argument is genuinely useful: it explains why the same handful of roles keeps reappearing in stories with nothing else in common.
The Russian who counted stories
The first person to make the problem tractable had no interest in psychology at all. Vladimir Propp, a Soviet philologist, published a slim and initially obscure study of Russian wonder tales in 1928. Rather than asking what the tales meant, he asked what they did, one step at a time, and discovered that beneath enormous surface variety the plots ran on a very small set of moves. His term for those moves was functions: an act defined by its consequence for the story rather than by who performs it or how. A hero is warned and disobeys the warning. A donor tests the hero. The tales differed wildly in costume and setting while agreeing, to a startling degree, on sequence.
The part relevant here is what Propp did next. Having isolated the functions, he grouped them into what he called spheres of action, each belonging to a role: the villain, the donor, the helper, the sought-for person and her father, the dispatcher, the hero, and the false hero. This is the crucial move and it is still misquoted. Propp did not claim there are seven kinds of person in stories. He claimed there are seven bundles of narrative work, and that any given character may pick up more than one bundle, or hand a bundle to somebody else halfway through. A character is a job description, not a personality. The second English edition of the Morphology has been the standard reference for structural folklore since 1968, and its central insight is still the cleanest definition of an archetype anyone has produced: a role identified by what it makes happen.
The pattern that swallowed everything
Two decades after Propp, Joseph Campbell made a much larger and much looser claim. The Hero with a Thousand Faces, first published in 1949, argued that myths worldwide are variants of a single underlying story, which Campbell called the monomyth, borrowing the word from James Joyce. The shape is departure, initiation and return: a protagonist leaves the familiar world, undergoes ordeals in a strange one, and comes back altered and carrying something the community needs.
Campbell was not the first to propose a hero pattern; Otto Rank and Lord Raglan had each published comparable schemes in the decades before him. What Campbell had was reach. He wrote for a general audience, and by the late twentieth century his scheme had become the default vocabulary of screenwriting manuals and studio notes, mostly through intermediaries who condensed it into a numbered checklist. It is now common to hear a story described as being at the threshold by people who have never opened the book.
What the folklorists actually objected to
Among the scholars whose material Campbell was summarising, his reputation is considerably worse. The objection is not snobbery about popularity, and it is worth stating precisely.
The complaint is about method. Campbell's universality claim is supported by examples selected because they fit, drawn from vastly different cultures, periods and genres, and stripped of the contexts that gave them meaning. Alan Dundes, one of the most prominent American folklorists of the last century, made this case bluntly in an address to the American Folklore Society later published in the Journal of American Folklore, arguing that the sweeping universalist claim collapses on contact with the actual distribution of tale types, and that motifs Campbell treated as global are simply absent from large parts of the world. Folklorists had already built a different tool: a catalogue of tale types that records where a story is actually found rather than assuming it is everywhere.
The fair conclusion is not that Campbell is worthless but that he is doing something other than what he said he was doing. He is not describing a law of human myth. He is describing a shape that a particular literary tradition finds satisfying, and describing it well enough that people have been building to that shape deliberately ever since. That is a self-fulfilling prophecy rather than a discovery, and it explains a great deal about why so many blockbuster films feel structurally identical.
The distinction Jung made and almost nobody keeps
Behind all of this sits Carl Jung, whose term archetype has been flattened almost beyond recovery. Popular usage treats archetypes as a set of stock characters, a deck of cards you can deal: the Mother, the Trickster, the Wise Old Man. Jung's actual position, set out across the essays collected in the ninth volume of his collected works, drew a line that the popular version erases entirely.
On one side is the archetype as such: not an image, not a content, but an empty formal tendency, a disposition to organise experience in a certain way. It cannot be depicted, because it is the predisposition rather than the picture. On the other side is the archetypal image: the actual figure that a given culture produces when that tendency meets local material. Jung was insistent that inherited ideas were not what he meant, and that the images differ everywhere while the tendency to produce some image or other does not. The trickster in one tradition and the trickster in another are not the same character wearing different hats; they are two culturally specific solutions to a similar pressure.
Whether those tendencies exist at all is a separate question his critics have pressed hard. The distinction is still worth keeping, because it names the thing that actually travels: a relationship, not a costume.
Why writers' rooms use them anyway
None of this scholarly weather much affects the working use of archetypes, which is embarrassingly practical. On a production with twenty people making decisions about the same character, an archetype is a compression format. Saying that someone is the sceptic, or the fixer, or the one who tells the truth at the wrong moment, communicates in four words what a character document takes four pages to convey.
The Cambridge Companion to Narrative's survey of character theory sets out the long argument between treating characters as people and treating them as textual functions, and working writers quietly occupy both positions at once. Early in development a character is a function, because the plot needs somebody to refuse the plan. Later the same character is a person, because an actor has to say the lines and no one can perform a function. Good rooms use the archetype to get started and then spend the rest of the process ruining it, since a character who never breaks the shape they were assigned reads as furniture.
Archetype is not stereotype
Which brings us to the confusion that does actual damage. Archetype and stereotype are not neighbours on a spectrum; they are different categories of claim.
An archetype specifies a role in a structure. It says what someone does in the story, and it is indifferent to who does it. The mentor is defined by giving the protagonist something they need before the ordeal, and anyone at all can fill it. A stereotype does the reverse. It attaches a fixed bundle of attributes to a category of person and asserts, in advance, what members of that category are like. One is a statement about narrative machinery. The other is a statement about people, and a false one.
The practical test is substitution. If you can change everything about a character except their function and the story still runs, you were using an archetype. If changing who they are breaks the joke or removes the point, you were relying on the audience's assumptions about a group, and you have written a stereotype however mythic the packaging. Propp's insight holds up rather well here, nearly a century on: a role is what it makes happen. Everything else is decoration, and decoration is where the actual character lives.
