The conflict on camera: The photographs that have defined the war in Iraq

It may yet be that the most memorable image of this war is one of the crudest and, what's more, one that only has its full effect as part of a moving sequence. The shot begins with flustered breathing and a canted shot of a man's face, with a white sky behind him. Then two circles of crimson appear ( see image, page 7), perfectly abstract shapes – until the camera is tilted downwards and the liquid begins to pool under the effect of gravity and trickle down. What for a second had looked almost like a symbolic representation of blood is suddenly recognisable as the real thing, as vividly present as if it had been dropped on to a microscope slide in front of us.

Click here for the Conflict on Camera gallery

It may yet be that the most memorable image of this war is one of the crudest and, what's more, one that only has its full effect as part of a moving sequence. The shot begins with flustered breathing and a canted shot of a man's face, with a white sky behind him. Then two circles of crimson appear , perfectly abstract shapes – until the camera is tilted downwards and the liquid begins to pool under the effect of gravity and trickle down. What for a second had looked almost like a symbolic representation of blood is suddenly recognisable as the real thing, as vividly present as if it had been dropped on to a microscope slide in front of us.

It should not be there, of course. It falls on a part of the camera that is always supposed to be transparent and invisible to us, and by doing so it suddenly and shockingly erases the secure distance between us and the events that we can dimly glimpse through the lens. This isn't just a television image, magically detached from what it records. It's the machine that made it, too – one with a vulnerable person behind it. As if to underline that fact, the cameraman reaches forwards with his free hand and tries to smear the blood away.

What followed, though, was curiously and significantly bloodless. John Simpson, the reporter, stumbles towards the camera, a small trickle of blood on his temple. His words conjure up a shambles. "There are bodies lying around," he says breathlessly, "there are bits of bodies on the ground." But we don't see that in this footage: and not simply because the attack is still in progress and the camera is pressed tight to the ground. We later learn that the team's translator has been fatally wounded in the legs and died of blood loss; that Simpson has seen men burn to death in front of him.

But we don't see images of those things either. The sentences are permissible, but the pictures are not. The rules of decorum – applied in northern Iraq or in the editing suites of Shepherd's Bush – spare us from the charnel-house reality. And oddly that makes those two drops of blood even more potent, because they have escaped the general circumspection about the literally visceral reality of explosive violence. As small as they are, they stand as representatives for so much that has been spilt unseen. In the hoariest cliché of all – they bring the war home to us.

That cliché has always had only limited truth when it comes to still pictures. From the very beginning, photography has made special claims for the veracity of its reportage, but it has also always been a protectively limited medium. Photographs don't smell and they don't shriek. What's more, as a visual form photographs are inextricably implicated in a larger visual culture, with its own history and traditions. When you see a photograph of a wounded soldier or a grieving woman it is all but impossible not to connect them – unconsciously or otherwise – with previous lamentations and depositions and pietàs. And since we assume that the subjects of such photographs are utterly sincere in their postures and actions, we naturally suspect the picture itself of having designs upon us. This is one source of the unshakeable anxiety about composition that war photographs stir in us – and one of the reasons why faults that would disqualify other kinds of picture from claims to greatness can often leave them enhanced. Robert Capa's blurred and inadequate photographs of the D-Day landings are exempt from the suspicions that will always linger around his formally beautiful photograph of a falling Republican soldier in the Spanish Civil War.

In another sense, though, the cliché about bringing the war home is truer now than it has ever been. Photographs these days don't just record the war for those who can't attend in person; they prosecute it by other means. The front page of a newspaper is a home front, a place where the battle for hearts and minds is fought. Many photographs – however pure the motives of those who took them – appear to have us in mind as a strategic target or as potential recruits. (This is not just a figure of speech, incidentally – British Army recruiting centres typically experience a surge of enquiries during hostilities because, however horrifying images of war are, almost all photographs implicitly confirm our conviction that we will be spectators of horrors, rather than their raw material.)

Sometimes the tactical imperative is brutally apparent, as when the Iraqi Ministry of Information released pictures of dead American marines to the world media. This was, in bluntly literal terms, an offensive gesture from a combatant that did not have many other means of making one. It was designed to wound and – in the term so freely used by Tony Blair during the war in Kosovo – to "degrade" the enemy's fighting capability.

But what made these pictures into an assault was less their content than their motive. During the Second World War, Franklin Roosevelt lifted an embargo on the publication of photographs of American fatalities to allow Life magazine to publish George Strock's now famous image of dead marines on a New Guinea beach. He believed that Americans at home had become detached from the gravity of the sacrifices being made overseas. The photograph shocked its audience, but looking at it was a kind of civic duty. Had it been distributed by the Japanese authorities, on the other hand, it would have been tantamount to treason to publish it and something like collaboration to grieve over its contents.

This war looks as if it will be too short to make such a patriotic memento mori necessary, and domestic opposition would make it hazardous anyway. But curiously Strock's picture has a kind of equivalent in Jon Mills's photograph of a dead Iraqi soldier in a trench ( page 8). As in Strock's picture no faces are visible, and as in Strock's picture the vital distinction between living flesh and dead earth is beginning to disappear. Strock's marines are already merging with the sand; this unknown soldier has his face pressed into the clay of his trench and only the hood of his coat makes it easy to distinguish his head from the clumps of earth around him. His posture reads as almost bashful, as if death is a shame he can't quite acknowledge, and above the parapet of what has now become a grave protrudes a white flag of surrender.

It was understandable that the widespread publication of this picture should lead to accusations of hypocrisy against some sections of the Western media. If images of our dead were unconscionable, how was it that their dead could be displayed like this? But the objection misses a crucial point about the picture, which is that its effect is largely self-questioning, not triumphant. How can we not read that white flag as a kind of accusation, for instance?

One of the most distinctive features of photography in recent wars is that images that would once have been the stock in trade of enemy propagandists – dead or injured children, shattered buildings, wailing mothers – are now firmly part of the repertoire of ordinary reporting. The celebrated photograph of an incinerated Iraqi soldier at the Mitla pass – one of the most famous images of the first Gulf War – was first published in a British newspaper, not an Iraqi one. Indeed you could argue that it is one of the conditions of democratic warfare that it is steadily deplored even as it is fought; and that photography provides one of the most effective means for doing that. Because photography is so insistent on particulars, it is a kind of universal acid against abstractions. You can talk in print of "fierce fighting" or "civilian casualties" – bland formulations that blur the impact on individuals – but photographs insistently return you to the component parts of such generalisations. Crucially, you can't take a picture of a good intention – only of the effects it leaves behind it, which often fall so far short of what might have been hoped.

It's not that triumphalism isn't possible any longer. The American raid on Saddam's Baghdad palace produced a classic instance of the photograph as boastful assertion. The raid itself was a rhetorical statement, a florid gesture of capability: and, since these forces did not intend to take the palace permanently, they had to take a picture to prove they'd actually been there. John Moore's image of US Army Staff Sergeant Chad Touchett ( page 7), sitting back in one of the palace's armchairs, is a potent image of invasion. This is surely how you occupy a seat of power; legs spread wide in an atavistic gesture of challenge, and surrounded by attendants and advisors. It is a knowing echo of Saddam's presence, a formal portrait of authority that is mockingly undercut by the contrast between the furniture and the clothes and the rubble strewn over the marble floors.

And, whether it knowingly replicates photographs of troops in Hitler's Reich Chancellery or not, it adds itself to a long roster of historic images of hostile intrusion. There are even simpler affirmations of power – images of the machines of war that draw on the dependable theatrics of gunfire. In Kuni Takahashi's picture of US marines firing artillery just south of Baghdad ( pages 4 and 5), the stark diagonals of gun barrels lifted over the low horizontal of a desert scene are unavoidably painterly. These are modern weapons, aimed by computer, but the combination of steel and smoke hasn't essentially changed since the First World War. (This image is, for all its compositional drama, essentially inert. It's difficult to tell, unless you're an expert, whether these are Iraqi weapons or American, and the knowledge wouldn't greatly affect how you feel about the picture's aesthetic virtues. It makes the point that while photographs need no translation, they often need some kind of subtitle.)

What has changed is how we represent the other end of the trajectory that arcs out from these muzzles. We no longer expect to see war without seeing its victims, intentional or unintentional. And this isn't a simple virtue, though it is often represented as that; as an uncomplicated increase in our honesty about war. Looking at such pictures, we learn that it is always civilians who are truly "embedded" in a war, unable to extricate themselves from this destructive incursion. The many photographs of civilian and military vehicles sharing the same roads outside Basra and other southern Iraqi towns made that point almost surreally. In other photographs the foreground offered you a picture of a morning commute and the background an almost medieval image of apocalypse, black smoke churning into the sky.

To look at such photographs is to perform due diligence, to get our accounts in order. This is the cost of war, and whether you're for it or against, it it is a moral obligation to include it in your calculations. The problem is that photographs of victims are as open to abuse as any other kind of photograph, both by those who disseminate them and those who look at them.

In a recent essay on the depiction of war, Susan Sontag cites an incident in the Balkans in which an identical photograph of children killed in a marketplace shelling was used by both sides as evidence of their opponents' heartlessness. Even when there's no dispute about what the picture represents, its meaning is not necessarily fixed. Looking at Ali Heider's powerful image of an Iraqi father weeping over the coffined bodies of his family, an opponent of the war is likely to feel a sense of useless, enraged vindication. This is what they said it would lead to. A viewer who felt that the war was a necessary evil might feel equally moved, but draw on the emotion to bolster a sense of grave endeavour.

Neither form of pity does the man himself much good; and it could be argued that, far from amplifying his reality to us, the photograph diminishes it. He becomes an instance of grief and the occasion, possibly, for a kind of soothing of the conscience rather than an affliction of it. It's hard to be certain, when we look at such images, that they don't reassure us about the continued operation of our sympathies. We prove that we care by looking at them, but the act is a kind of ritual that barely detains us from our daily lives. The care with which we now audit the deficits and losses of war hasn't made us notably less likely to fight them.

It is common to talk of photographs as "defining", as if they dictate the terms on which we view a war. It can certainly feel like that retrospectively. Looking at Larry Burrows's powerful images of wounded soldiers in Vietnam, you feel that they offer a perfect image of a country enmired. Robert Capa's blurred images of the D-Day landings seem like an ideal representation of determined thrust. But the truth is that the war also defines the photographs, as well as the other way round. And this war – so variously imagined and so disputed in its nature – hasn't yet delivered a really canonical image.

The most likely candidates for preservation, I would guess, are pictures of smoke in the distance. The slanting plumes of oil fires, gracing an empty horizon, have been a feature in countless shots, often with a soldier in the near foreground inspecting the emptiness ahead of him. But we've seen similar pictures before, from the Gulf War, and the Six Day War, and even Suez. The one genuine novelty of this conflict has been the astonishing son et lumière of night-time bombing of Baghdad – not viewed from above, through the green-tinted murk of a bomb camera, but photographed from across the river by journalists. In many of the images, street lights shine beneath the mushroom clouds and floodlights sculpt their undersides. The combination is an unprecedented one: a city in the relative tranquillity of night, not concealing itself in any way from attack, and yet marked by this localised plume of intense destructive power.

To have taken such a picture in most previous wars would have required extraordinary recklessness. In this one it just needed a tripod and patience. Where all the other pictures of this war have emphasised its continuity with previous battles – with the hard slog of fighting on foot and the messy spill of violence beyond its intended targets – these photographs show us something. What they don't tell us is what to think of it. Whether the precision that they depict is profoundly sinister or admonitory is still being fought over in the streets of Baghdad.