I was never squeamish. But I wasn't particularly drawn to the bloody or grotesque, either, so it's odd to think that I've now devoted 20 years of my life to investigating violent death as a forensic pathologist in New York City and Florida.

After medical school in London, I moved to the US and trained in pathology. And I found it incredibly dull.

Don't get me wrong: surgical pathology is vitally important, highly skilled work. But with medical examiner cases, to borrow an expression from Damon Runyon, "a story goes with it". And I loved forensics immediately.

I flew down to Miami to apply for a training position. The chief, Joe Davis, drove me through the city in his big maroon Cadillac.

He was an exceptional tour guide, a connoisseur of the city's mayhem, quick to point out the site of a decapitation here or a bombing there.

That year, in the medical examiner's autopsy room in Miami, I got an education in how we live, as seen through the fish-eye lens of how we die. It was hallucinatory and intense.

Catapulted out of the placid gentility of life in Boston, I suddenly found myself squatting over a mangled body, explaining to sweaty cops that the victim had been hacked to death with a machete, not hit by a truck, as they were claiming.

Becoming a forensic pathologist felt like stepping into a role on a TV show. I don't know whether detectives talk like detectives because that's how they talk, or because they've learnt from TV how detectives talk.

Expectations for the role of "forensic pathologist" are not as clearly sketched out; I made it up as I went along.

I moved to Miami Beach.

It was an absurd life. I'd find myself driving home at 4am after examining a body dumped out in the swamps of the Everglades, flashing my badge at the highway toll booth as my tiny car boomed out the latest bass sounds.

Then over the causeway, into Miami Beach, where blonde models in bikinis rollerbladed down Ocean Drive in the moonlight. Even as I lived it, my life felt unreal.

My work, on the other hand, was all too real. The city had one of the highest homicide rates in the country, as the cocaine wars ended and the crack epidemic began.

Adventure

Cocaine poured into the country through Miami; we saw many cocaine mules, poor South Americans who'd swallow over a hundred packets of cocaine then board a flight to the US, only to die in some downtown fleapit when a packet ruptured inside them.

For someone coming up in the field, it was an amazing experience. I was young then, and the whole thing -- crime scenes, murder, the autopsy room, working with the cops, testifying in court -- had a delirious sense of Hollywood adventure.

I was emotionally detached from the work, which is critical: you can't let yourself get busted up by every death you investigate. My work felt like watching a film.

Forensic pathologists are gatekeepers of death, chronicling the human experience from a highly specialised perspective. For a long time, I thought of death as something that happened to other people -- it seemed like that should be part of the bargain.

It wasn't until after 9/11 that I finally realised I'd been mistaken all along -- death wasn't an abstraction, and no one in my line of work gets left unscathed.

In the carnage of a violent crime scene, it's easy to lose sight of just how perfect fluid blood is.

A blood droplet falls through the air just like a drop of water, which is to say as a perfect sphere -- the teardrop shape artists use to depict rain is wrong.

A skilled forensic scientist reads the blood spatter around a murder victim as if the droplets were word bubbles floating above a comic strip head.

Shape and size are critical: blood falling vertically onto a smooth surface leaves a circular dot, but with uneven surfaces, the shape is a sunburst, little spikes of blood splashing away from the centre.

If the bleeding person is running, the bloodstains have an elongated shape, like teardrops pointing in the direction of movement. From the drop's proportions, criminalists can calculate the angle at which it struck; strings or lasers are used to recreate droplet trajectories; and if there are enough drops, the point of origin of the blood spray can be identified.

In practice, you can usually get a rough sense of how the assault unfolded based on quick pattern recognition -- the groupings of the droplets, their location, their height above the floor, and so on.

Blood from a severed artery creates a distinctive pattern of rivulets on a surface, while pale ghost outlines may reflect someone standing between the victim and the wall, absorbing spatter.

There is, of course, a learning curve.

At one of my first murder scenes, the victim, whose throat had been cut, lay stretched in front of the TV. I knelt to examine the body and within seconds my legs were soaked with blood hidden in the dark carpet.

I also quickly learnt to look up when entering a bloody death scene. In a stabbing or beating, blood from the weapon can be flicked up onto the ceiling (this is particularly common with hammers, which have considerable angular velocity).

Gunshot wounds spray out a superfine mist of blood; if the victim is close enough to the gun, blood can be recovered from the barrel, which makes it extremely useful in this age of DNA.

The Holy Grail of crime scene evidence is the killer's fingerprint in the victim's blood.

Visually, murder scenes are often extremely dramatic -- police always overestimate the amount of blood at a scene because even a small amount (a cupful, say), spilled over bathroom tiles, makes the place look like a slaughterhouse.

Bloody crime scenes play well on the big screen, upping the stakes by underscoring the reality of the violence.

Movie pathologists were once shadowy figures whose only function was to confirm the hero's brilliant deductions, usually while chewing a sandwich over an open cadaver.

But the recent vogue for forensics has produced characters like Dr Randall Frazier (Erik Dellums) in The Wire. As with everything on the series, the part of the medical examiner was realistically written, particularly in the early seasons.

Some criminologists are often derisive about forensics in movies and TV, but I like it.

Shows like CSI Miami aren't meant to be forensic science: they're forensic science fiction, distilling the coolest parts of criminalistics to their coolest essence.

That said, I like crime fiction to be accurate; after you've seen a killer's pathetic attempt to write in his victim's blood at a crime scene, you just can't accept a book where the psycho covers the walls with vast screeds in blood.

In one serial killer case in which I was involved, the murderer used to leave after strangling his victims until they stopped moving.

Movie and TV strangulations last just a few seconds, but in real life a fatal strangulation takes a long time: with my case, the killer was caught when two of his victims regained consciousness after he'd left, and described him to the cops.

No matter how it looks in the movies, strangling people is hard work.

Jonathan Hayes is a forensic pathologist in New York City, and the author of two novels, Precious Blood and A Hard Death

- Jonathan Hayes