Four years ago, I tried to see how using this blog might spur a regular writing schedule. It worked pretty well in the summer of 2013, and then the new school year came along, and the blogging stopped.

But the reading and writing and collecting didn’t stop, and now it’s time to try to gather the material in a new form. In the near future, that means that I will be *withdrawing* certain posts from this blog, in order to rewrite them in a new format. As the work progresses, I expect to make some new posts, too—especially bibliographies and suggestions for others who might be interested in the project of “thinking through images.”

As it has evolved, this work has much less to do with the history of cinema, or even with the work of Godard and others. Instead, it points in the direction of a general pedagogy of words and images, aimed at exploring more fully what we are capable of thinking. Even after all these years of very slow progress, this task seems more urgent than ever.



Here’s a very simple illustration of montage from La Chinoise (1967):

chinoise montage

Léaud is talking to the camera. He is looking down, and when he lifts his eyes, the shot cuts to Wiazemsky, sitting still and looking at the camera, then another cut returns us to Léaud looking down.

Then back to Wiazemsky, looking to the left, who starts to turn right—then the movement is cut and resumed by Léaud, who finishes turning all the way right.

The logic of the sequence is perfectly clear: static alternation is combined with dynamic continuity.


The sequence continues as Léaud proceeds with his monologue, cut together with an array of photos (Brecht, Shakespeare, Mayakovsky) and reverse-shot interplay with the cameraman and sound technician. It’s an extended textbook demonstration of elementary montage technique. (We will leave aside the question of whether it is also a textbook example of  Brechtian Verfremdung and/or Lacanian suture.)


I’ve arranged each short sequence as if it were a line of poetry on a page: in this way, it becomes possible to see the rhythms and accenting of the sequence at a glance. Each line has its own montage “device.” (Note what happens in the fourth line: Léaud’s face is suddenly lit up, as if from a flashbulb. You don’t have to stop the shot to give it a shock.)

“For my part, I have made only one discovery in cinema:

how to pass flexibly from one shot to another, between two different movements,

or even what’s more difficult, from a moving shot to an immobile one.

It’s something that almost nobody does because they never think of it;

it’s a matter quite simply of taking up the movement at the stage where the previous image left it.

So one can link any image to any other,

for example an automobile to a bicycle

or a crocodile to an apple.”

—Jean-Luc Godard, interview (1967)

This is indeed a momentous discovery: “one can link any image to any other.” Not only does Godard discover that this linkage is possible in principle; he claims to have found a reliable way to make it happen in practice. Such is the path of all great discoveries: we learn what is possible by figuring out how to do it, and from that point on nothing looks the same. If the Bresson quotation (in the previous post) told us that montage is impossible when each image stands alone, this quotation tells us that montage is always possible, as long as you know how to make the images go together.

Rather than treating this discovery as if it were just an idiosyncratic tweak of the rules of film editing—as if everything is understood as soon as we call it a “jump cut”— it would be better to approach it as an immense theoretical revolution.

Let’s look at the examples. It seems easy enough to combine an image of an automobile with an image of a bicycle. Beyond the immediate impression that they “belong” together, we can invent narrative, scenic, or logical motivations for connecting them. In fact we can make connections according to many different criteria: we see links between images on the grounds of similarities and contrasts, causes and effects, sequences or simultaneities, scales, genres, styles, colors, and so on. Indeed there may be countless reasons to connect images, and yet (common sense tells us) not all images can be connected.

What about the crocodile and the apple? What could possibly connect them? There seem to be only a few ways to go: either we rack our brains in search of a reason for the link (do crocodiles eat apples? is there a story that unites them? are they symbols of nature? etc.) or we invoke a poetic link between the objects (do they “resemble” or “stand for” each other in some hitherto unexpected way?). In a pinch, even the apparent disjunction between them could provide a motive for connecting them, as though Godard were making some kind of surrealist joke.

All of our brain-racking will be in vain, however, because it misses the main point. If there is to be a connection between the crocodile and the apple, it will not be a relationship but a dynamic; it will operate at the level of the specific images at hand, and not at the level of crocodiles or apples in general. In fact it doesn’t really have to do with crocodiles or apples as such, but only with discovering the possible movements that might pass through them and the trains of thought that might involve them. Montage-thinking is a matter of finding the elements in each image where a movement can be carried onward, “flexibly,” without reducing one to the terms of the other. Godard wants to make montage “only from what is there in the image, from the signifying and not the signified.” It is a serial movement of sense rather than a punctual production of meaning. It makes its own frames of reference, always more or less open, sometimes along many axes, sometimes along just one. That is why it may be too much to say that there is actually a “connection” between the images: the movement of montage “undoes” each of the images in the process, remaking both by creating the gap that separates them.

A few years before, Godard described this complex dynamic in a perfect phrase:

“Two shots which follow each other do not necessarily follow each other.

The same goes for two shots which do not follow each other.”

Now, circling back to our starting point, we can say that the same dynamic must have been at work with images of automobiles and bicycles, too: it must be possible to combine those images as if we did not already know what they are and what they mean. (Bresson again: if you already know what an image means, it will be useless for cinematic thinking.) Here is a new creative task: how is it possible to see cars and bikes without resorting to generic images, as if they are as strange to each other as crocodiles and apples?

And once we’ve used the “obscure” pair of images to open up the “obvious” pair, we should try new combinations: what about automobiles and crocodiles? Or apples and bicycles? Why not? In each case, we can ask: what movements, what thoughts, can be invented here? Do apples “go with” bicycles better than automobiles? How would you make sense of a crocodile and a bicycle? This exercise is not some attempt to return to Lautréamont’s “chance encounter of a sewing machine and an umbrella on an operating table.” What we can call “the rule of the crocodile and the apple” does not require metaphoric ingenuity or surrealist extravagance; it is not a matter of constructing a logical category or narrative thread to connect them from outside. It is a matter of learning to think about things in terms of movements, to the point where even the terms “thing” and “movement” seem too vague to designate what is actually going on in each series of images.

Montage pursues sense in the most rigorous, most concrete way possible, and in discovering how to use montage to see and to think this way, Godard has acted more like a research scientist than a lyric poet.


The main quotation comes from “Lutter sur deux fronts,” Cahiers du cinéma, October 1967; reprinted in Godard par Godard, volume 1. (The quotation appears on page 307.) The second quotation comes from “Pierrot mon ami,” Cahiers du cinéma, October 1965; also in Godard par Godard, volume 1, page 263.