> ## Content Index
> Fetch the complete content index at: https://photopedagogy.ghost.io/llms.txt
> Use this file to discover other available public pages before exploring further.

# From the Bookshelf
- URL: https://photopedagogy.ghost.io/bookshelf-fontcuberta/
- Published: 2026-04-21T17:20:00.000Z
- Updated: 2026-09-18T17:23:51.000Z
- Description: Joan Fontcuberta — Against Barthes: The Eye and the Index
- Author: Jon Nicholls
- Tags: Bookshelf

Joan Fontcuberta begins his book with an accident. When he was about 13, a science experiment went wrong and he lost part of his index finger. He doesn’t show you the injury or dwell on it. But it stays with you as you read, because the index finger turns out to be the book’s central character, the pointing finger, the finger of evidence, the finger that says: look here, this happened, I was there. And Fontcuberta’s is incomplete.\* This is a book about photography’s claim to show us the truth of things, and it is written by someone who carries, literally, a reminder that such claims are always provisional.

![](https://substackcdn.com/image/fetch/$s_!tmAp!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4b8007c1-8f0b-4e8c-8172-5cb25125decc_2297x3598.jpeg)

Photographs open and close the book, drawn apparently from the archive that runs through it. On the cover and page 9, a young man points at his black eye. The final image in the book shows a man in a doorway pointing at his chest, head bowed, slightly melancholic. Both images appear to come from a Mexican crime tabloid called [Alerta](https://willstraw.com/mexico-imitators-of-alarma-1980s-and-1990s/?ref=photopedagogy.ghost.io), active from the 1960s to the 1980s, whose thousands of photographs Fontcuberta initially encountered through a curator at the Musée Niépce. But these two images feel different from the others. In every other photograph in the archive, the pointing finger is directed outward, at a wall, a tree, a patch of ground or the scene of something that happened to someone else. Here, the finger points inward. The subject is pointing at himself. The correspondent, the witness, the victim and the crime scene have all collapsed into one person.

![](https://storage.ghost.io/c/cc/10/cc10dd61-f9df-448c-99a4-1ae3f24596e3/content/images/2026/09/5bf32d63-03f6-4140-b999-64fd1503071a_2488x3369.jpg)

![](https://storage.ghost.io/c/cc/10/cc10dd61-f9df-448c-99a4-1ae3f24596e3/content/images/2026/09/ed6d8f31-aa36-4acd-a182-47a155a0f36a_2188x3176.jpg)

[Fontcuberta](https://www.scienceandmediamuseum.org.uk/what-was-on/joan-fontcuberta-stranger-fiction?ref=photopedagogy.ghost.io) has spent his career inserting fictions into documentary archives, fabricating scientists, inventing space missions, creating species that never existed and presenting them as photographic evidence. If anyone would smuggle himself into a Mexican crime tabloid to make a philosophical point, it is him. Could they be AI generated self-portraits? It’s a tantalising prospect but I can’t verify it. In a book whose entire argument is about the unreliability of photographic evidence, that unverifiability feels appropriately unresolved.†

Nota Roja (red news) is a Mexican tradition of sensationalised crime reporting, graphic and unsparing. The publication Alerta was one example. Whilst moving through thousands of its archive images, Fontcuberta noticed a recurring motif. In image after image, someone stands at the scene and points. A finger extended toward a wall, a tree, a patch of ground or a doorway. Here. This is where it happened. Look at this. The pointing finger is photography’s most literal attempt to do what language does more easily, to locate meaning, to pin the image to a specific claim about the world. And yet, as Fontcuberta shows, the pointing alone is never enough.

![](https://storage.ghost.io/c/cc/10/cc10dd61-f9df-448c-99a4-1ae3f24596e3/content/images/2026/09/417e9035-9bfe-4e1f-ad8e-8c3ab3446901_2580x1873.jpg)

![](https://storage.ghost.io/c/cc/10/cc10dd61-f9df-448c-99a4-1ae3f24596e3/content/images/2026/09/dfb16862-4ea2-4d59-b5c6-40963a5c8d96_2689x1895.jpg)

A young woman stands beside a tree, arm extended, index finger aimed at the trunk. What does the image tell us? That she was there. That the tree was there. That something brought a photographer to this spot. But what? She might be pointing at her favourite tree, or indicating which one needs cutting down, or showing a friend something she noticed on the bark. The image is not empty. It is full of possible meanings and open to multiple readings. What it is not is self-evident. The caption on the back of the print provides the context. “Young Teresa Valdivia shows our correspondent the place where she was raped by her father, the porter Cipriano Alemán”.

![](https://storage.ghost.io/c/cc/10/cc10dd61-f9df-448c-99a4-1ae3f24596e3/content/images/2026/09/42999896-96b7-4fb8-98ef-41ee94c2180c_3466x2280.jpg)

![](https://storage.ghost.io/c/cc/10/cc10dd61-f9df-448c-99a4-1ae3f24596e3/content/images/2026/09/afeb7cd3-dbb7-424d-b6af-0ebd65f58848_2066x3563.jpg)

Teresa Valdivia

Only with the words does the image acquire its particular and devastating weight. The pointing finger, which might have meant anything, suddenly means one specific, unbearable thing.

Fontcuberta’s point is that this dependence on language is not a peculiarity of crime photography. It is the condition of **all photographs**. We tend to assume that a photograph shows us what was there before interpretation begins; that it denotes first and only then do we start to read it. But pure denotation, he argues, is a theoretical convenience rather than a description of how photographs actually work.

> *Everything is connotative and the only differentiating factor is the intensity with which a thing connotes*.

Before he deals with Barthes directly, Fontcuberta discusses the guillotine. He reminds us that the first assistant at an execution, known as ‘the photographer’, was responsible for ensuring the condemned person’s head was correctly positioned for the descending blade. This required physically handling the face, tilting it into position. In the slang of the time, *tirer le portrait*. To pull the portrait. And the crouching posture this required reminded onlookers of a studio portraitist, ducked beneath a black cloth, arranging their subject. “The executioner doesn’t kill people, the guillotine does,” Fontcuberta writes. “By the same token it isn’t the photographer but the camera that extracts life from you to turn you into a spectre.”

This is Fontcuberta at his most characteristically playful and vertiginous, finding in an obscure historical detail a connection that illuminates his argument. The portrait photographer and the executioner share a gesture. The camera and the guillotine share a function. Photography, from its very origins, has been entangled with extraction, with fixing, with a kind of violence done to the living in the name of memory.

![](https://storage.ghost.io/c/cc/10/cc10dd61-f9df-448c-99a4-1ae3f24596e3/content/images/2026/09/64139378-9009-4d0e-b693-869a00637ae7_3021x3021.jpg)

![](https://storage.ghost.io/c/cc/10/cc10dd61-f9df-448c-99a4-1ae3f24596e3/content/images/2026/09/ac32a695-1289-42e2-90f6-30e1e1c29d2c_3021x3021.jpg)

Which brings us to Roland Barthes, whose 1980 book [*Camera Lucida*](https://en.wikipedia.org/wiki/Camera%5FLucida%5F%28book%29?ref=photopedagogy.ghost.io) is both Fontcuberta’s starting point and his sparring partner. Barthes proposed that every photograph carries an irreducible truth claim. Whatever else it might mean, whatever interpretations it generates or resists, the photograph insists that something was there. The light bounced off a real thing, passed through a real lens, left a real trace. He called this the *ça-a-été* (that-has-been) and he felt it with particular force while going through photographs of his mother after her death. He found it in a picture taken when she was a young girl, before he was born, standing in a winter garden. The photograph proved she had existed in a way nothing else quite could. But, Fontcuberta asks, if the ça-a-été is photography’s essential truth claim (something was there), then **what exactly was there?** To make the point he turns to the great catastrophic images of the twentieth century.

> *Captain Fortino Sámano, cigar clamped between his lips, facing a firing squad: ça-a-été. The Hindenburg blazing in the sky: ça-a-été. A Republican militiaman falling at Cerro Muriano: ça-a-été.*

And so the list continues, through Iwo Jima, Hiroshima, photographs of lynchings, Dallas and Tiananmen and the Twin Towers and Bin Laden’s assassination: *ça-a-été*, *ça-a-été*, *ça-a-été*, *ça-a-été*, the phrase accumulating until it becomes something close to an incantation that is simultaneously a recognition of photography’s extraordinary power and a gentle destabilisation of it. But *what has-been* is not straightforward. Some of these images are contested. Some are staged. Some have been misidentified, mis-captioned or decontextualised. This is Fontcuberta’s quarrel with Barthes, or at least part of it. He doesn’t deny the indexical trace, the physical connection between the image and whatever stood before the lens. It’s just that this trace is insufficient to settle the meaning of the image.

There is a further layer to the argument that photographers and photography teachers may find helpful, because it concerns the theatricality that photography tends to pretend it doesn’t have. The Alerta photographs were staged. The witnesses were carefully positioned. Pointing fingers were directed. The photographers arranged the scene before releasing the shutter. This is not unusual. Fontcuberta reinforces the point through a precise comparison between cinema and photography:

> *Consider how we read a film and how we read a photograph. It seems clear that film actors avoid looking at the camera because doing so will break the fourth wall, reminding us of the presence of the film crew and spoiling the illusion of a transparent, direct, unmediated representation. The actors should play their part as if there were no cameras, lights, sets, or technicians involved. When an actor directs their gaze at the camera, the intrusion of artifice is revealed, and we are stripped of our privileged status as observers. In a film, for credibility not to be lost, the conditions in which the images are manufactured must be invisible. In a photograph, on the other hand, when the subject looks at the camera that act implies a reciprocal recognition of presence(s), a dual awareness of contact. In cinema, staging is camouflaged to pretend at naturalness. In photography, the staging is not hidden and is so obvious that we forget about it — we do not see it — precisely because it is so visible.*

Only involuntary photography, the surveillance camera or accidental exposure, escapes theatricality entirely. Everything else involves choices, arrangements, framings, decisions about where to stand and when to press the shutter. The photograph’s claim to have captured the real is always a little more complicated than it looks.

![](https://substackcdn.com/image/fetch/$s_!XVK5!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd57ee300-8ff4-4118-b229-648f7bd9bf2a_2801x2016.jpeg)

What makes all of this more than an academic argument about semiotics is what happens when Fontcuberta follows it to its logical conclusion: Artificial Intelligence.

He dates the moment when photography’s truth claim became technically optional with characteristic precision. 1989\. Photoshop was set for release.‡ The Berlin Wall fell, history ended and photography celebrated its 150th anniversary in what he describes as “a rapture of navel-gazing, utterly oblivious to the impact of the colossal meteorite that was to wipe from the surface of the earth the photographic ancien régime and its tyranny of reference.”§ From that moment, the chain of physical causation that underwrote the indexical promise, light bouncing off something real, leaving a trace, could be severed at will.

But this rupture didn’t appear out of nowhere. Fontcuberta traces photography’s deep entanglement with the cultures of accumulation, archiving and control that preceded it.

> *From an economic and political perspective, photography contributed to the domination and control of the world through its symbolic appropriation, and its development went hand in hand with the colonial expansion of the great powers: the possessing of the world in images was the preamble to its subsequent physical possession. Wherever the photographer set up his tripod, an army planted its flag.*

If photography gave birth to a culture of accumulation, of archives, statistics, surveillance and the ordering and classifying of the world, then AI, with its vast training datasets and its culture of prediction, is its logical heir. This time, however:

> *The images will be born without strain, they will be brought into the world by algorithmic midwives trained to birth perfect babies every time.*

AI images, Fontcuberta argues, no longer emanate from the real but rather from text. Born from language rather than from the world. Generated from descriptions and statistical patterns, they arrive without a referent, without anything having stood before any lens. The camera’s role as witness is made redundant. But this is not, for Fontcuberta, simply a catastrophe. What AI makes explicit is something that was always true and always concealed. The photograph was never just recording reality. It was constructing a possible version of it. He suggests thinking of *fiction-images*, pictures made with Photoshop, or generated by AI, or assembled from fragments of other images, “as something like an act of image-making whose outcome is the simulation of fragments of possible worlds, in which simulation is an authentic degree of reality and not a denial of reality as such.” An image that generates real responses, real emotions, real consequences in the world has a kind of reality regardless of whether it records something that existed. The fictional and the documentary are not opposites. They are different points on a continuous spectrum, and the camera was never quite as close to the documentary end as we liked to believe.

For photography educators, this presents an interesting challenge. The question of what an image is, how it works, what claims it makes and whether those claims can be trusted, is not a new question introduced by AI. It is the question photography has been asking since Niépce pointed a camera out of an upstairs window. Fontcuberta quotes [Moholy-Nagy’s argument](https://photoquotes.com/author/laszlo-moholy-nagy?ref=photopedagogy.ghost.io) that the illiterate of the future would be the person who could not read photographs, then extends it:

> *our greatest ignorance, the most crucial illiteracy today, is our inability to keep up with computing, robotics, and artificial intelligence.*

This is not a specialist skill for technologists. It is the condition of critical engagement with a world saturated in images whose origins are increasingly opaque.

I find this argument persuasive and urgent for photography education.

Gert Biesta, the educational philosopher whose thinking runs through much of what I write here, argues that the [fundamental gesture of teaching is pointing](https://fotologic.substack.com/p/the-only-way-is-ethics-part-2). Not explaining, not delivering knowledge. Pointing. “Hey you, look at this!” The teacher redirects the student’s attention toward the world without controlling what the student sees or concludes. It is an open gesture. It does not force or guarantee. It simply says: **here**. Fontcuberta’s index finger is the same gesture. Photography points. Teaching points. The camera says look at this and so does the teacher, and neither can determine what the looking will find. **The pointing is the pedagogy**.

What makes this particularly resonant in Fontcuberta’s case is the half-finger, the philosophy of provisionality written into his own body by a failed experiment. The pointing is incomplete. The certainty is missing at the tip. The teacher who points with absolute certainty has already foreclosed the encounter. The photograph may claim to show, definitively, what was there but this is misleading. Educators who respond to AI-generated imagery by refusing to engage with it are, I think, making a version of this mistake. They are pulling the caption off the Teresa Valdivia photograph and insisting the image is now meaningless. But the image was never only its caption. And to refuse to point at something difficult is not a neutral act. It is a choice.

---

Thanks for reading PhotoPedagogy! Subscribe for free to receive new posts.

[Subscribe](#/portal/signup/free)

*These posts will always be free but, if you enjoy reading them, you can support my analogue photography habit, and that of my students, by contributing to the film fund. Thanks to those of you who have already done so. All donations of whatever size are very gratefully received.*

[Buy me some film](https://buymeacoffee.com/photopedagogy?ref=photopedagogy.ghost.io)

---

\* Fontcuberta himself invokes the figure of Thomas, the doubter who required physical evidence, who had to put his finger into the wound before he could believe. It's a telling choice of patron saint for a career spent questioning photography's claim to show us the truth of things. The difference, as Fontcuberta might put it, is that his finger is itself the wound.

† I have contacted someone at MACK. I’ll update the text to confirm or deny my suspicions if I get a clarifying response from them.

‡ [https://en.wikipedia.org/wiki/Adobe\_Photoshop](https://en.wikipedia.org/wiki/Adobe%5FPhotoshop?ref=photopedagogy.ghost.io)

§ [https://en.wikipedia.org/wiki/The\_End\_of\_History\_and\_the\_Last\_Man](https://en.wikipedia.org/wiki/The%5FEnd%5Fof%5FHistory%5Fand%5Fthe%5FLast%5FMan?ref=photopedagogy.ghost.io)