Contribution
Folktale study before Propp classified tales by category (Miller, Wundt) or by plot type (Aarne’s index; Volkov’s fifteen plots) and described them by motifs (Veselovsky) or by Bédier’s constant “elements”. Propp argues that each of these cuts across the material, because the parts of tales move freely between tales. He proposes a new unit, the function of a character, and shows on Afanasyev’s tales 50–151 that these units are few, ordered, and common to all wondertales. After the book, a tale can be described as a sequence of symbols, two tales compared by their schemes, and the class of wondertales defined by structure rather than content.
Key insight
Ask what the characters do, not who they are. Baba Yaga, Morozko, a bear and a wood-spirit all test the stepdaughter and reward her: one function, four performers. Functions are “extremely few, and characters extremely many”, which explains at once the tale’s dazzling variety and its monotony.
Assumptions
- The material. Afanasyev’s collection, nos. 50–151 by his own numbering, taken as given rather than chosen; justified by the claim that functions repeat so much that a hundred tales exhaust them, with a larger “control” material looked at but not shown.
- The class. “Wondertales” are provisionally Aarne types 300–749. By ch. IX the class is redefined by the result: “a tale built on the regular alternation of the functions given, in various forms, with some absent from each tale and others repeated”. Tales outside the scheme are by that definition not wondertales.
- Function identity is fixed by consequence. Identical acts are different functions if they lead to different things (a test followed by a magical agent is the donor’s test; followed by marriage, a difficult task; followed by a quest, a dispatch). The analyst decides by what follows.
- Authenticity. Pure structure is “proper only to the peasantry, and a peasantry little touched by civilisation”; Grimm’s tales show it less purely; deviations in humorous tales count as decomposition.
Key results
- Thesis I. Functions of characters are the constant, stable elements of the tale, independent of who performs them and how.
- Thesis II. The number of functions known to the wondertale is limited: 31, after an initial situation (α).
- Thesis III. The sequence of functions is always the same. Ch. IX qualifies it: inverted sequences occur (D–G–Z before A; departure before learning of the misfortune); struggle can follow pursuit; recognition and exposure, wedding and punishment, can swap; transfiguration is the least fixed. These are called oscillations, “not a new compositional system”.
- Thesis IV. All wondertales are of one type. Every function can stand in one tale with every other, except struggle–victory (B–П) and difficult task–solution (З–Р), which occur in 41 and 33 of the 100 tales and together in one move only three times; in two-move tales the struggle move comes first and the task move second, and the two-move tale is “the basic type of all tales”.
- Spheres of action. Seven: villain (villainy, fight, pursuit); donor (preparation and provision of the magical agent); helper (spatial transference, liquidation, rescue, solution of tasks, transfiguration); princess and her father (tasks, branding, exposure, recognition, punishment, wedding); dispatcher; hero (departure, reaction to the donor, wedding); false hero. Hence “the tale knows seven characters”.
- Helpers. Universal (only the horse, in this material), partial and specific (one function: the ball of thread guides, the sword kills); “the magical agent is nothing but a particular form of the magical helper”.
- Linked variants. Some paired functions bind their subtypes absolutely (a fight in the open field with victory in the open field); some partly (abduction with several kinds of recovery); the rest combine freely, and “it is by replacing one subtype by another of the same element that the metamorphosis of tales proceeds”.
- Generation. “One can create new plots artificially in unlimited number”: take any A, one of the possible B, then C↑, any D, then G, and so on, omit or treble elements, and assign functions to characters; “the schemes come alive, they become tales” (answering Shklovsky’s “by laws not yet known”).
- The teller’s freedom. Not free in the order of functions, in linked subtypes, sometimes in the attributes a function requires, and in what the initial situation must include; free in which functions to omit, how each is realised, the naming and attributes of characters, and language.
- Plot and variant. Predicates give the composition and subjects and objects the plot; distinguishing plot from variant is impossible (“from qualitative the question becomes quantitative”), so the whole stock of wondertales is “a chain of variants”.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | In this corpus, the constant elements are characters’ functions, independent of performer and manner | strong (for the corpus) | ch. III, with examples from each tale cited; the schemes in the appendix (not checked here) |
| C2 | There are 31 functions, and every tale’s action falls within them, apart from a few “unclear elements” | moderate | ch. III; the material is 100 tales and the control material is not shown |
| C3 | The order of functions is invariant, up to named inversions and swaps | moderate | ch. IX; the qualifications are Propp’s own |
| C4 | All wondertales are of one compositional type | moderate, partly definitional | ch. IX; the class is redefined by conformity to the scheme |
| C5 | Functions distribute into seven spheres of action, i.e. seven roles | moderate | ch. VI |
| C6 | Tales can be generated artificially from the scheme | weak | asserted as an experimental check; no generated tales are shown |
| C7 | All wondertales derive by transformation from the dragon-abduction tale | weak | ch. VIII–IX; “a very bold claim, since we do not give the picture of transformations in this work” |
| C8 | The wondertale is morphologically a myth | weak | ch. VIII–IX, offered as a hypothesis for the historian |
Note of 2026-10-20: the appendices were read in a second text, the HTML of the 1969 Russian edition at gumer.info. The per-tale schemes still could not be checked, because Appendix III says the schemes are printed “на вклейке”, on a fold-out insert, and the HTML does not carry it. The appendix does fix the sample behind C2 and C3. Of the fifty tales whose schemes were tabulated, five were left out: No. 94 as a tale of another kind, and Nos. 123, 127, 139 and 159 because their moves are interrupted too often and “the canonical sequence is also not strictly kept”. Propp also says that the material he printed is half his working material. So the order invariance is shown on 45 printed schemes, chosen after setting aside the tales where the order was not strictly kept. That makes C3 an order claim about a filtered set, which is a reason to narrow it further than “the hundred tales”. Appendix II gives full analyses of a few tales (including Nos. 131 and 247), and one of its notes says some tales break the canon through parodic treatment. The 1968 English edition at the Internet Archive is lending-restricted and was not read.
Method
Compare tales across plots; isolate the constant actions by abstracting from performer and manner; name each by a noun of action with a symbol and subtypes; fix each by its consequence, which resolves assimilations (identical acts serving as different functions) and double meanings (one act serving two functions at once); then transcribe each tale as a string of symbols, separate its moves (from each villainy or lack to its resolution), and stack the strings to read off the order and the alternatives. Non-function elements are catalogued separately: connectives (how one character learns what another did), trebling devices, motivations (“the most unstable” elements, probably late), forms of entry of characters, and attributes.
Concepts
- функция (function): an act of a character defined by its significance for the course of the action.
- ход (move): a development from villainy (A) or lack (a) through intermediate functions to a wedding or another resolution; each new villainy or lack opens a new move.
- круг действий (sphere of action): a group of functions belonging logically to one role.
- ассимиляция (assimilation): one function realised in the form proper to another.
- начальная ситуация (initial situation): the opening enumeration of family or hero; a morphological element, not a function.
- недостача (lack): the morphological equivalent of villainy, “comparable to zero, which in a row of figures has a definite value”.
- детонация (“detonation”): a mismatch between a move’s opening and its resolution (Ivan sets out for a horse and returns with a princess), material for the study of transformations.
Connections
Goethe’s morphology frames the book (the epigraphs on metamorphosis and the Urpflanze). The critics engaged are Veselovsky (motif and plot), Bédier (constants and variables), Aarne (the type index), Volkov, Wundt and Shklovsky. The book defers two things to later work: the transformations of tales (named here as a separate study) and the historical roots of the wondertale. Doja, Capocchi and Santucci (LIT-804) list Propp’s morphology among the narrative formalisations computational work has attempted. Lévi-Strauss’s structural analysis of myth (LIT-775, unread) is the usual counterpoint; the Italian edition (1966 by Dundes’s dating in the 1968 English edition, NOTE-642; this note first gave 1969) carried his intervention and Propp’s reply (per Open Library), not read here.
Bearing on the record
- Produces THEORY-167: the functional composition of the wondertale, with what it does not establish.
- Transformation and transmission. Propp’s account of what a teller can and cannot change is a claim about which features of a tale survive retelling: the order of functions and linked subtypes are constrained; performers, attributes, motivations and wording are free, and “everything that enters the tale from outside submits to its norms” (a devil becomes villain, helper or donor). That is a structural thesis about the corpus, not a measurement of retellings; the record’s transmission-chain readings (LIT-769, LIT-771, LIT-820) are the kind of evidence that could test it.
- Saussure. Propp’s constant-and-variable analysis is independent of Saussure (LIT-774), whom the OCR text never names; his model is Goethe’s morphology, not linguistics.
- No instruction for machine-learning practice; nothing for the anthology.
Limitations
- One collection, Russian, nineteenth-century, edited by Afanasyev; a hundred tales, from a starting point fixed by Afanasyev’s numbering.
- The coding is Propp’s alone; nothing in the book tests whether another analyst would assign the same functions, and assignment by consequence gives the analyst latitude.
- The universality of the order is stated with exceptions that are then explained away (inversions, oscillations, humorous decomposition).
- The one-type thesis is helped by defining the class as the tales that fit the scheme.
- The study of attributes and of transformations, which the preface says the full version contained, was cut; claims resting on them (the basic form, the myth hypothesis) are asserted without the evidence.
Open questions
- Does an independent coding, by several analysts with agreement measured, of wondertales outside Propp’s hundred reproduce the 31 functions and their order?
- Do the B–П and З–Р branches and their order hold in other traditions, or only in the Russian?
- If tales are generated from the scheme as Propp proposes, do listeners accept them as tales of the kind, and does retelling move non-conforming tales toward the scheme?