Narratives: Stories and Structures

on 2 December 2008 at about 20:59

This item was originally posted on CEMP's Interactive Media Portal on 02 December, 2008.

We continue our journey through story by looking at the internal structure of narratives.

Notes from the second keynote in the Narratives series. Post comments or use the forum if you want to clarify anything! Here’s last week’s introductory lecture. TL;DR? This week’s shoutometer – or suggest next week’s – or even write a story based on one of George Polti’s 36 situations.

We see the whole structure, and the man in the tree. We can see structure, but not texture. The man in the tree sees the texture of the tree, he is in close and intimate. But the man in the tree cannot see the whole structure of the tree, unless he leaves the tree. We cannot be in the tree and outside it at the same time. We can’t see texture and structure at the same time.

Is structure all-important?

Man With a Movie Camera: The Global Remake – here’s an interesting project which takles Dziga Vertov’s 1929 film Man with a Movie Camera and allows anyone to upload video clips as alternatives to Vertov’s original. A new ‘remake’ is made using software and the database of clips that people upload.

The project is very tightly structured around each shot from the film. In theory, the original and the remake should be structurally identical. So does this mean that the two films are the same?


The Stories and Structures lecture focussed on structuralism. Structuralism is the movement in intellectual thought which developed over the course of the C20th, mostly out of the work of Russian formalists like Vladimir Propp and the work of the Swiss linguist Ferdinand de Saussure.

Saussure’s structuralist approach to linguistics led to what we now call semiotics, and is the study of the internal structure of language. He called a synchronic snapshot of language ‘langue’ and distinguished it from ‘parole’ – the diachronic utterances which ‘langue’ facilitates.

‘Langue’ is relationally structured and it is this relational structure which allows words to ‘have’ meaning. Every word is in a paradigmatic relation to every other word, and these paradigmatic relationships enable and define difference and meaning – different words, different meanings… This sounds complicated, but isn’t really: if we had only one word, we could only ‘mean’ one thing. We have lots of different words, and differences between them (the relational structure) are what create meaning.

Claude Levi-Strauss

Claude Levi-Strauss was the main player in this lecture. Levi-Strauss did to mythologies what Saussure did to language: looked for a relational structure in the myths and legends of various cultures. In his essay, ‘The Structural Study of Myth’, Levi-Strauss analyses the Oedipal myth. Interestingly, he emphasises that he feels free to take many different instances of the myth from throughout the ages, and combine them into the material to be analysed.

So Levi-Strauss’ snapshot of the mythic narrative is synchronic like Saussure’s analysis of language, but not in the sense that he examines a myth at one moment in time, but rather that he treats the entire myth as a system which can be laid out at a moment in time and surveyed as a paradigmatically and syntagmatically relational structure. He does exactly this, in the form of a table-like grid, into which he places the key components – or mythemes – of the story. He repeats this system in other works, such as those which treat native american myth systems. Again, his synchronic ‘snapshot’ of myth is not confined to one story, but to the entire collection of stories which a particular culture keeps in circulation.

The primary tool Levi-Strauss uses to define the components, the mythemes, which go into his table, are ‘binary oppositions’.

Binary oppositions

One of the keys to understanding structuralism is the notion of ‘dyads’ or ‘binary oppositions’. It may help to think of binary oppositions as fundamental categories of human thought. If we are born into a world of undifferentiated chaos (babies have to learn to make sense of the kaleidoscope of sensory information that bombards them from the moment of birth), then the development of mind in a human being must involve categorisation. In order to define something, you must also define what it is not. Every new definition can be further split into yet more categories, ad pretty much infinitum.

In the lecture I compared this to mitosis. For those of you yearning for more philosophical insight, if human analysis is rather like a knife, which differentiates the world into conceptually different things, then this is what we might call the origins of ‘dialectic’.

So anyway, pairs of such categories might include:

These binary oppositions tend to get paired up with another binary opposition:

So raw is bad, cooked is good (if you don’t want to die of food poisoning); kin is good if you want people to co-operate with; but kin is bad when you want to reproduce and make healthy babies.

The structural law of myth

In the essay on the Oedipal myth, Levi-Strauss laments that various accidents in the field of anthropology have led to the undermining of prospects for the ‘scientific study of religion’. He goes on to say that analysing the instances of the Oedipal myth, what is aimed for is a ‘logical treatment of the whole’ that will lead to uncovering the ‘structural law of the myth’.

These descriptions of his programme start to outline one key aspect of what structuralists want to do: approach the object of their study scientifically. Levi-Strauss emphasises this by noting that when poetry is translated from one language to another, serious distortions of the original poem occur, but translations of myths do not suffer in the same way. He describes myth as ‘timeless’, and as something that functions at a ‘high level’, which ‘takes off’ from language. These are characteristics which make myths translatable, reproducible: they have ‘constituent units’ which are governed by laws in just the same way that particles have laws which are governed by the laws of physics.

Roland Barthes say something very similar in his essay, ‘Introduction to the Structural Analysis of Narrative’ when he describes narrative as ‘international, transhistorical, transcultural’ – he even uses the U-word: ‘universal’.

Reductionist approaches to human culture

The logical conclusion of the emphasis on structural units which are universal, reproducible, predictable, translatable, international, transhistorical and transcultural is to to take a reductionist approach to human culture (see the notes on structure and texture from the last lecture). Indeed, it is the inevitable consequence of adopting a scientific method at all, which is certainly what structuralism attempts.

A classic method of structuralist anthropology, for instance, is to identify the ‘component units’ of human behaviour which can be observed in every known human culture. These include the telling of stories, taboos on incest, music, co-operation, war and hundreds of other ‘universals’.

This inevitably raises questions about human nature itself: if all human societies have wars, then is war a ‘natural’ trait of humans, etc? In some senses these questions are really very useful and we can the pursuit of knowledge addressing difficult ethical issues. However, we might also want to make some pretty severe criticisms of such ‘universalist’ approaches to human nature. It is one thing to link incest taboos to biological imperatives. It is another to impose ‘laws’ on human society because of particular interpretations of narratives such as religion and myth.

Scientism and Marxism

We will return to deal with Marxism and criticisms of scientism in future lectures. For the purposes of this lecture, we simply note that Marxists such as Adorno and Horkheimer of the Frankfurt School regarded scientific, universalist approaches to culture as tyrannical; there are many ways in which we can note that rationalism contains self-contradictory problems (do something as simple as look for the square root of 2 and you have found what is called an ‘irrational’ number). But we end up confronted by a ‘representational fallacy’ – we can think that we are dealing with concrete reality, when in fact we are dealing with nothing more than mist conjured by our imagination: language, story, myth, narrative.


So much for the theoretical matters we were trying to nail down. We also tried to illustrate them. First of all, let’s think about Memento (Nolan, 2000). SPOILER ALERT!!! This film illustrates the difference between story and plot rather well. If someone asked you ‘what happens in Memento‘, you could well give two very different answers, both equally correct.

One of these is the story – the chronological story in which Leonard Shelby is an insurance claim investigator, whose wife is raped and murdered, and who ends up with anterograde amnesia, and repeatedly seeks out vengeance by tracking down and killing people called John G – either because other people set him up to kill them for their own purposes (Teddy setting Lenny up to kill Jimmy), or because Lenny himself sets himself up to kill them out of spite (killing John ‘Teddy’ Gammel) because that is better than admitting that he has become ‘no-one’, with nothing to live for.

The other answer is the plot – what we ‘experience’ as we watch the film. Lenny Shelby kills a man, and we gradually learn through a series of flashbacks that Lenny is seeking to avenge his murdered wife. He learn that he uses tattoos and polaroid photographs to keep a record of what he knows, so that he can piece his life back together every time he ‘wakes’ into a present with no context. He stresses that he goes on ‘facts’ – but as each scene unfolds, our trust in Lenny’s ability to record facts erodes, until we realise that the so-called facts that he uses to direct his actions are actually little more than the subjective expressions of momentary emotions (such as his trust of Natalie) or even the result of his own malice (his mistrust and murder of Teddy). In the final scene of the film, we discover not only that Lenny has killed at least two ‘John Gs’ before killing Teddy, but we are also provoked into worrying that the entire ‘story’ we have just witnessed may have been an entire fabrication: that Lenny is in fact, Sammy Jankis; that he unwittingly killed his wife using repeated insulin injections; and that he, not Sammy Jankis, has been institutionalised, and that we may be watching nothing more than a phantasm, a momentary hallucination.

Structuralism vs formalism

Structure and form sound like very similar things. In some contexts, they can even mean the same things. But they are quite different things when it comes to narrative theory, so let’s try to pull them apart a little.

I suggested a massive simplification here, which we may use provisionally to better understand structuralism and formalism. Let us think of structuralism as interested in story, the totality of the diegesis. Formalism on the other hand is interested in plot – the framing devices through which we learn the story.

I emphasise that this is an exaggerated simplification, simply to help us get the idea.

narrative as formal system

So structuralism wants to map out the whole of the diegetic story – the fabula, rather than the sjuzet, and understand its structure. Story can be plotted in many different ways, and this need not alter the story. Levi-Strauss even says that we can use the many different instances of the Oedipal myths together to understand its ‘laws’.

So to the extent that this simplification is useful, formalism is interested in the artifice of representation – that is the devices that we use to tell these stories – the form they take (which is quite different from the ‘structure’ of the story). We’ll be placing our emphasis more on formalism in the next lecture.

Analysing stories

We created an example of a story by using a fairly common structure which we then dressed in the texture of your own narratives. The structure took the form: a protagonist; who encounters and obstacle; then overcomes the obstacle; and there is a positive outcome. Your stories looked like this:

A science teacher is trying to teach a class. A monster enters the class and makes teaching rather difficult. The teacher destroys the monster using their well-established and well-founded scientific principles and understanding. And the class can continue as normal.

A penguin wants to go into the sea. Unfortunately there is an apple in the way. So the penguin kicks the apple, which skittles away, leaving the way open for the penguin to get to the sea.

A student has crippling dyslexia; he is helped by his teacher, enabling him to write a novel which makes his fortune and he becomes extremely rich.

An alcoholic cowboy wants to drink himself to death. He walks out of a bar, and encounters someone who gets in his way. He shoots them, and is able to carry on his way, and drink himself to death.

All of these stories can be expressed in terms of a simple binary opposition: complete / incomplete. The teacher is incomplete, because he is unable to fulfil their role of teacher; destroying the monster allows them to perform their role, thus achieving ‘completeness’. The penguin wants to feed in its natural milieu, but is unable to do so – it is incomplete; it achieves completeness by kicking the apple and getting to the water. The student is incomplete, because he is unable to write his novel; the teacher’s help complete him, by allowing him to fulfil his dream. The alcoholic cowboy is incomplete because he desires to kill himself, but his way is barred; being free to do so makes him complete.

Being able to see stories in this way gives us a way to analyse structure of the narrative. Sometimes it is very easy, as in Hollywood blockbusters, when there are easily identifiable good guys and bad guys. Such obviously binary structures are Manichean, as was George Bush’s famous line, ‘you’re either with us or against us’.

Limits of structure

While these structuralist ideas are undoubtedly appealing and powerful in many ways, they also have shortcomings. We will return to how structuralism becomes post-structuralism in the final lecture of the series.

For now I just want to think about how Memento parallels the rational approach to making sense of the world that structuralism attempts. We watch Lenny relying on things that seem to be very clearly facts. Polaroids document facts. Lenny tattoos ‘facts’ on his body, and as we begin our journey with him, we trust him – we feel sure he must be a just avenger of his wife’s murderer. But this fact-based approach deteriorates into wilful self-deception.

In fact, it we can even read Memento as having no story at all: the chronological story we witnessed becomes indeterminate, as we see one event multiple times (such as Lenny pinching his wife leg, but later the same event is shown as Lenny injecting his wife’s leg with insulin). What we thought was the internally coherent diegesis of Memento is really a hall of mirrors, where we cannot be sure that anything we have seen is not an illusion, a hallucination. Memento has become pure plot – a trick or an artifice, an illusion created through the device of story-telling.

Next time

Next time in Narratives, we’ll turn to formalism, and examine the artifice of plot. We’ll take in some Marxism, and we’ll also consider mimesis.

Invitation to a structuralist experiment

George Polti described 36 dramatic situations. If we write lots of stories based on the situations, we can analyse them to see if the structuralists are right. Here’s a website where you can add a story based on any of the 36 situations. We’ll look at the stories there ate the end of teh Narratives series and see what sort of conclusions we can draw…

Too long; didn’t read

An important school of theoretical thought in the C20th wanted to systematise cultural study. This was the structuralist movement, and it aimed at adopting a scientific approach to the analysis of myths, narratives and language.

Structuralism has some strengths and some weaknesses: narratives certainly have internal structures, and structuralism certainly helps us to understand how language works; however, whether it is a good idea to think of narrative structures as producing ‘universal’ human laws is a different question entirely.

[Original] Comments

I like structuralism, i like saussure’s idea that language influences the way that you see the world, as without language how do we have a universal understanding of experiences, even though people who speak a different language experience different things to us due to language…

I like it too, I recall my first introduction to dyads when I was studying the history of the novel at university, and it rather blew my mind (that, the Front242 tape and the acid).

But I don’t like universalism. I’m not sure we do have universal understandings of experiences, do we? We don’t get to be outside of the tree, looking in. So rather than universal understandings, I like to imagine we have lots of overlapping venn diagrams of intersubjectivity.

Lecture archive

  • Media & The Body Readings
  • The Use of Theory
  • Play
  • Production - [Critical Media Concepts and Contexts]
  • Narratives: Endlessnesses and Existence
  • Narratives: Endings, Meaning and Morals
  • Narratives: Performers and Players
  • Narratives: Performers and Players
  • Narratives: Familiarity and Strangeness
  • Intro to Media and Participation 2008 - 2009
  • Narratives: Stories and Structures
  • Narratives: Opening and Introductions
  • The Writerly Text: Part 1
  • Media & Participation: Identity
  • Media & Participation: Truth
  • Media & Participation: Citizenship
  • Media & Participation: Culture
  • Ownership of Ideas: Part 2: The History of Copyright
  • Ownership of Ideas: Part 1: The Romantic Author
  • Bournemouth Soundseeing: collaborative authorship
  • Marx's Critique of Capital: 101
  • Key concepts: Ideologies ...a historical view
  • Narrative and Structuralism and the brothers Grimm
  • Intro to Digital Media at BU
  • Small Print

    Joe Flintham 2000 - 2015

    Long form: Menticulture

    Professional Services: Fathom Point