You are here : Home Page / Theories / Greimas / The Canonical Narrative Schema

Available Languages

Follow us on

SignoSemio's Facebook SignoSemio's Twitter

The Canonical Narrative Schema

By Louis Hébert

Professor, Université du Québec à Rimouski



Algirdas Julien Greimas

Using the canonical narrative schema (CNS) we can organize the elements of an action into a structure consisting of five components: (1) The action component can be broken down into two components itself: (2) competence, which results from the factors that are required in order to accomplish the action (wanting-to-do, having-to-do, knowing-how-to-do, and being-able-to-do) and (3) performance, the actual realization of the action, made possible by the acquisition of competence. (4) Manipulation is the component that deals specifically with wanting-to-do and having-to-do. (5) The last component, sanction, has to do with evaluating whether the action was truly realized, and the corresponding retribution (reward or punishment) that the performing subject has incurred. Here is an example of an action based on the CNS: The King asks (manipulation: having-to-do) the Prince to rescue the Princess (action). The Prince trains for combat (competence: knowing-how-to-do and being-able-to-do) and then rescues the Princess (performance). The King then grants him (sanction: positive retribution (reward)) half of his kingdom and the hand of the Princess.

This text can be found in extended version in this book:
Louis Hébert, Dispositifs pour l'analyse des textes et des images, Limoges, Presses de l'Université de Limoges, 2007.

Click here to obtain the English translation of this book.

This text may be reproduced for non-commercial purposes, provided the complete reference is given:
Louis Hébert (2006), « The Canonical Narrative Schema », in Louis Hébert (dir.), Signo [online], Rimouski (Quebec),

An updated and extended version of this chapter can be found in Louis Hébert, An Introduction to Applied Semiotics: Tools for Text and Image Analysis (Routledge, 2019,



First we will present the canonical narrative schema (CNS) in summary. (For this chapter, we shall draw liberally, but not strictly, from Courtés 1991, 98-136.)

Using the canonical narrative schema, we can describe the logical, temporal and semantic arrangement of the elements of an action (which may or may not be represented as narrative programs (NPs)) using a structure that has five components: (1) The action, which is broken down into two components itself: (2) competence (which has to do with the necessary prerequisites of the action: wanting-to-do, having-to-do, knowing-how-to-do, and being-able-to-do) and (3) performance (which concerns the actual realization of the action); (4) manipulation (the component that deals specifically with wanting-to-do and having-to-do); and (5) sanction (which has to do with evaluating the action and the retribution it entails (reward or punishment).

The CNS is "canonical" in that it adequately accounts for the general organization of action of a large number of narrative productions (texts, movies, etc.).


Earlier we mentioned logical, temporal and semantic organization. Now we shall give some details. The components are bound together by presuppositional relationships. For example, the sanction (the presupposing term) presupposes the action (the presupposed term). We will come back to these relationships. These presuppositions generally form the basis for relationships of temporal succession (with the presupposed term being temporally anterior to the presupposing term). Finally, each component is part of a class of action, and because of this, each element is assigned a specific semantic label according to where it falls (an element of manipulation, an element of sanction, etc.).


Since we will often refer to NPs, we will quickly summarize what they are. (See the chapter on narrative programs for more details.) The narrative program is an abstract formula used to represent an action. The short formula of the conjunctive narrative program is: PN = F {S1 — (S2 n O)} and that of the disjunctive narrative program: PN = F {S1 — (S2 u O)}. S1 is the subject of doing; S2 is the subject of state; O is the object; n is the conjunction (with the object) and u the disjunction (lacking the object) between the subject of state and the object.

For example, in the fable "the Crow and the Fox", we have the following (conjunctive) narrative program: NP = F {Fox — (Fox n Cheese)}. In simpler notation, the narrative program is written as follows: S1 — S2 n O, for example, Fox —Fox n Cheese.


The canonical narrative schema, proposed by Greimas, takes the theoretical place of the actantial model, also developed by Greimas. Let us look at the primary differences between them.

  1. The actantial model revolves around a subject and an object. If we reduce this pair to a NP, we observe that, for one thing, the subject and the object are implicitly linked by a junction (conjunction or disjunction), and for another, this triad corresponds to the second state of a NP. As for the CNS, it centres explicitly on a NP.
  2. As compared with the actantial model, a pair of actants has been dropped: the helper/opponent pair. The helping and hindering elements are integrated within competence, and if they are considered explicitly as actants, it would only be in the context of a NP, which is composed solely of subject and object actants. Thus, the object of a NP of competence is an element of the competence, for instance, the acquisition of knowing-how-to-do. A helper is then a subject of doing in a narrative program of maintenance or acquisition of competence; an opponent is a subject of doing in a narrative program dealing with loss or non-acquisition of competence.
  3. Two kinds of senders are distinguished in the CNS: the sender-manipulator (of the manipulation component) and the sender-judge (of the sanction component). The receiver, known as the receiver-subject, corresponds to the subject of doing (S1) of the schema's central NP, where it is depicted in its relationship with one of the two senders.

Some methodological considerations: Equal attention may be given to each constituent element of the CNS, or selected elements may be chosen as the focus instead. In the latter case, secondary constituent elements should usually be analyzed at least briefly. As with the actantial model, we must first situate the central element (action) by careful selection, since all of the other elements depend on this central element. Once we have selected the action, we should make sure that the manipulation, sanction and competence we are describing actually do apply to this NP, and not some other NP in the same story. Consider the narrative sequence: PN1: Thief n Stethoscope, PN2: Thief n Contents of Safe, PN3: Thief n Prison. If we situate NP1 at the centre of the schema, we cannot have NP3 as the sanction, since the latter is the sanction for NP2, and not NP1.


The following diagram is a visual representation of the CNS (a slight modification of the representation in Courtés 1991, 100).

The arrows indicate presuppositional relationships between components; for example, the sanction presupposes the action, but the action does not presuppose the action.

Representation of the canonical narrative schema
Canonical narrative schema


We will specify the various presuppositional relationships between components as we go along, but at this point, let us mention a few logical and temporal principles. As one would expect, a presupposing element is optional relative to its presupposed element(s). This is the case, for example, with the sanction (presupposing) as opposed to the action (presupposed). A presupposing element, which is logically anterior, is usually temporally anterior, but not always.

We will give an example of temporal simultaneity between components that are in a presuppositional relationship: unless there are extenuating circumstances, wanting to blink one's eyes (competence) and doing it are perfectly simultaneous. Next we have an example of inversion between temporal arrangement and logical arrangement: if compensation in full for services is payable in advance, then the retribution stage of the sanction precedes the action at least partially. (It is not inconceivable that cognitive retribution (recognition) or pragmatic retribution in the form of an unanticipated bonus (money) could possibly follow the action.)



Manipulation (a term that has no pejorative connotation in semiotics) is the component of the CNS that has to do with changes in wanting-to-do and/or having-to-do. Positive manipulation aims to produce, increase or maintain them if they are already at adequate levels; negative manipulation aims to destroy, decrease, or maintain them if they are at inadequate levels. The purpose of positive manipulation is to cause doing; the purpose of negative manipulation is to cause not-doing.


The sender-manipulator directs his/her manipulation towards the receiver-subject, that is, the subject who is supposed to accomplish or not accomplish the action. In the NP representing this action, the receiver-subject corresponds to the subject of doing, abbreviated as S2. Positive manipulation is represented thusly in a NP: PN = F1 [S1 — F2 {S2 — (S3 n O)}]; In this formula, which differs from the standard narrative program (there is an additional subject, and S1 and S2 become S2 and S3, respectively), S1 represents the sender-manipulator, and S2 the receiver-subject.


By means of the explicit or implicit contract it establishes between the sender-manipulator and the receiver-subject, manipulation draws the action to be accomplished or not accomplished into the realm of possibility, along with the positive or negative retribution that will ensue if the contract is fulfilled or not fulfilled.


Take the case of two opposing armies, each one commanded by a general. The general of the first army (the sender-manipulator) uses positive manipulation (encouragement, the prospect of earning medals, threats, etc.) to incite his soldiers (the receiver-subjects) to advance (or at least not retreat) and he uses negative manipulation (threats, explosions, etc.) to cause the enemy not to advance (or even to retreat). The general of the second army (the anti-sender-manipulator) directs positive manipulation toward his soldiers (the anti-receiver-subjects or receiver-anti-subjects) aimed at the same action (to advance or at least not retreat).


The action is the central component (conceptually and visually) of the CNS. This action is generally represented by a narrative program. The action component is broken down into two components itself: competence and performance.

Action (or more accurately, performance) presupposes manipulation. (If there is action, then there had to be manipulation, but manipulation, even when successful, does not necessarily lead to action, as the element of being-able-to-do could be insufficient, for instance.)


Theoretically, in order for a NP to fit into the canonical narrative schema, the NP must be a performance NP, that is, a reflexive action, where S1 = S2 (for instance, washing oneself), as opposed to a transitive action, where S1 ≠ S2 (for instance, washing someone else). In practice, we are not sure that this principle is necessary.

The other components of the CNS may each be described using a NP (see our application below).


Competence (a term with no meliorative connotation in semiotics) is the component of the CNS that has to do with changes (creation, maintenance, increase, decrease, loss) in the prerequisite elements of performance (accomplishing the action).


There are four different modalities involved in competence: two that are also factors in manipulation- wanting-to-do (abbreviated as w) and having-to-do (abbreviated h )- and two others- knowing-how-to-do (k) and being-able-to-do (a).


When speaking of competence or one of its modalities, we say it is positive when it is sufficient to lead to performance; otherwise we call it negative. Negative competence or a negative modality has a minus sign in its notation (for instance, -w signifies negative wanting-to-do). In order to have performance, competence must be positive for both knowing-how-to-do and being-able-to-do, firstly, and secondly, for wanting-to-do and/or having-to-do. (Positive wanting-to-do can compensate for negative having-to-do and vice-versa.)

In contrast to what Courtés says, it appears that all positive competence leads without fail to performance; if not, then competence was not completely or truly positive. (For instance, I start to lift my arm, but a meteorite kills me. By appearances, I had competence, as in the general competence required to raise my arm, but it was not true competence suitable for that specific action in those specific circumstances.) Considered in this light, it follows that the relationship between competence and performance is more accurately a relationship of reciprocal presupposition: If there is (positive) competence, there will necessarily be performance; if there is performance, then there was necessarily (positive) competence.


Performance is the component of the CNS that has to do with the action's realization (in the strict sense), which was made possible by positive competence. Performance presupposes competence (and manipulation, naturally, since it involves wanting-to-do and having-to-do, just as competence does): If there is performance, then competence was necessarily positive. And we have mentioned that this presupposition can be considered reciprocal: when competence is truly and fully positive, realization of the action necessarily follows.


Performance, and therefore competence as well, is categorical and/or incremental. For example, going off a cliff is generally seen as a categorical action: one either succeeds or one doesn't. (A half-success or near-success is still a failure.) An election is an example of performance that is both categorical and incremental: victory is factual first, by obtaining at least 50% of the ballots cast plus one; but the incremental dimension should not be ignored: The intensity of the victory increases the closer we come to obtaining 100% of the votes cast.


Sanction is the component of the CNS that concerns the epistemic judgment (evaluation) of performance and the accompanying retribution that the performing subject has incurred.


The sender-judge directs his sanction toward the receiver-subject, that is, the subject who was supposed to accomplish or not accomplish the action (the subject of doing, S1 in the NP).


Epistemic judgement determines whether performance conforms to the implicit or explicit contract that was made during the manipulation stage. We must answer questions like the following: Was the action realized, and properly so? Is the presumed receiver-subject the right one, an impostor, or a case of mistaken identity?


Retribution is the next stage. It can be categorical or incremental, positive (reward) or negative (punishment), and pragmatic (gold, for instance) or cognitive (recognition, for example).

Retribution presupposes epistemic judgment (but not the reverse, since the sender-judge could die before giving the promised reward, for instance). Sanction presupposes action (or more accurately, performance that took place or should have taken place), but the action does not necessarily presuppose a sanction (as would be the case if the sender-judge died before passing his epistemic judgment, to take our previous example).


* * *

"The Crow and the Fox"
Jean de La Fontaine

At the top of a tree perched Master Crow;
In his beak he was holding a cheese.
Drawn by the smell, Master Fox spoke, below.
The words, more or less, were these:
Hey, now, Sir Crow! Good day, good day!
How very handsome you do look, how grandly distingué!
No lie, if those songs you sing
Match the plumage of your wing,
You're the phoenix of these woods, our choice."
Hearing this, the Crow was all rapture and wonder.
To show off his handsome voice,
He opened beak wide and let go of his plunder.
The Fox snapped it up and then said, "My Good Sir ,
Learn that each flatterer
Lives at the cost of those who heed.
This lesson is well worth the cheese, indeed."
The Crow, ashamed and sick,
Swore, a bit late, not to fall again for that trick.

* * *

In La Fontaine's "The Crow and the Fox", we will consider the action wherein the crow (C) sings at the fox's (F) urging. Here is a simplified CNS that describes this action.

A canonical narrative schema in "The Crow and the Fox"
  • F ? (C ? C n Song)
    This NP corresponds to:
  • F ? C n Wanting-to-do
    This NP presupposes:
  • F ? C n Possibility of positive retribution

(Cognitive) retribution of false contract

  • F ? C n Glory
C ? C n Song

Epistemic judgment

  • F ? Action n Positive Judgment
    (The action is judged as being properly realized.)


Pragmatic retribution: the real contract

  • F ? C u Cheese(dispossession)

Cognitive retribution: the real contract

  • F ? C n Humiliation
  • R ? C n Lesson (ironic retribution)
  • F ? C n Wanting-to-do
  • C ? C n Knowing-how-to-do (action already realized)
  • C ? C n Being-able-to-do (action already realized)


  • C ? C n Song (action realized)

A few explanatory details are in order.


During the manipulation stage, the fox implicitly offers two contracts, the real one and the false one. The crow believes that in return for singing, he will receive positive cognitive retribution in the form of glory from showing off his beautiful voice. Keeping the cheese does not even enter into the false contract, since singing and keeping the cheese are not mutually exclusive from the crow's perspective, or at least he forgets their incompatibility momentarily under the influence of the fox's flattery. The real, implicit contract is the following: If he sings, the crow will receive negative retribution, both pragmatic (the loss of his cheese) and cognitive (humiliation).

The fox concedes that the crow is beautiful at the outset, but says he wants to hear a voice that he assumes is beautiful ("No lie, if those songs you sing / Match the plumage of your wing..."). The fox's intellectual superiority is emphasized by the hyperbole of his flattery, which reveals the crow's gullibility and vanity - a crow is neither a peacock nor a nightingale. However, we should point out that the stupidity attributed to the crow is undeserved, for in reality, the crows are very intelligent animals (for example, it has been demonstrated that they can count to four).


During the sanction stage, the fox becomes the judge-manipulator. The epistemic judgment does not present any problems: The action was indeed realized according to the contract (the real one as well as the false one). However, for the retribution, the fox dispenses with pretence and grants the retribution provided by the real contract, to which the crow was an unwitting and unconscious party. He cynically idealizes the crow's loss as an exchange: After all, didn't the crow receive a lesson for his cheese? It is a very unequal exchange, in which a pragmatic object that is a basic necessity is exchanged for a cognitive object that is supposedly unique and not to be shared, although the fox does not deprive himself in giving it - he retains the lesson and, more importantly, the intelligence and the ruse of which the lesson is simply a manifestation.


COURTÉS, J., Analyse sémiotique du discours. De l'énoncé à l'énonciation, Paris: Hachette, 1991.
SPECTOR, Norman B., ed., The Complete Fables of Jean de la Fontaine, Chicago: Northwestern University Press, 1988.


Formulate a CNS from the main action in The Lord of the Rings by Tolkien, wherein the Hobbit Frodo Baggins destroys the evil ring of power.

Share this page