Dog Detectorists

i-7b128852736d83dbe8f4c804dd5a6691-dognose2.jpgAt the request of Aard regular and archaeologist Mathias Blobel in Freiburg, Germany, here's a summary of a recent paper in Swedish.

In Fornvännen 2007:3, husband & wife historians of archaeology Drs Ãsa Gillberg and Ola W. Jensen note that there are currently ongoing attempts to train dogs to sniff out archaeology, much like they have proved capable of finding illegal drugs, recently buried murder victims, even cancerous tumours. Results so far are inconclusive. The point of Gillberg & Jensen's note is instead to draw attention to an obscure Swedish antiquarian, Johan Lindman, who made the same suggestion already in 1761 in a Gothenburg newspaper.

"...might not the sense of smell guide people or animals in the search for [buried] money?" Everyone knows that copper coins and other metal objects smell, and so no man should

"... doubt that they could smell coins and metals, and that the vapours touch their noses when they run across buried money. It would be pedantry to want to prove such an evident truth. All I want to add pertains to the vapours of the metals, and the way to train the dogs. The vapours will be strongest when the metals become softened, as it were, by heat or moisture. This will be the time for treasure hunting. Surfaces give off vapours, but a hoard of buried coins has many surfaces, and so a softened lump of buried coin will smell stronger than the same mass of uncoined metal. Likewise, the smell of softened metal that has been buried at the same spot for a long time will be much stronger than that of natural pieces of metal that have never been touched. Thus will hidden treasure doubtlessly impregnate the upper layers of earth with rising particles, damaging to human health, and piercing to the nose of a dog."

Lindman had recently heard of a Gothenburg dog that had found a brass ring for his owner, but he had not performed any experiments of his own. Nevertheless, he made several suggestions on how to train treasure-hunting dogs.

He suggests an early start. The puppy's owner should let it play only with money, placing coins around the house and letting the dog search for them. The dog should be trained to mark finds by barking, but to keep to precious metals and ignore rock, wood and fabric. Lindman feels that the dog would need to concentrate on sniffing, and so should be made to wear a blindfold or a cap to avoid distractions.

Whether anyone ever took up Johan Lindman's suggestions and trained themselves a dog detector is not known.


Gillberg, Ã. & Jensen, O.W. 2007. Arkeohunden förr och nu. Fornvännen 2007:3. KVHAA. Stockholm.
[More blog entries about , , , ; , , , .]

More like this

The more dog training changes, the more it stays the same. I would love to have seen the 18th century style cap the money-sniffing pup would have been made to wear.

Thanks for taking up my suggestion, Martin.
I think this is an interesting concept, something that can be used alongside the more traditional geophysical methods as well as on parts of a site that weren't supposed to be excavated anyway or even for stray finds in the spoil heap.
Apparently dogs have recently been used successfuly in archaeology (for example in Prague: http://www.k9forensic.org/prague-zlicin.htm )
Those interested can visit http://k9forensics.org/historical.html for more examples. These dogs were of course trained to find bones which means you can probably use traditional law enforcement-trained animals instead of having to train them specifically for archaeology.
Nevertheless I'm pretty sure the process outlined by Johan Lindman would work. The problem I see with it is that you could only train the dog to recognize one specific smell (or at least very few smells) meaning that you would have to know in advance what it is you were trying to find. You'd also have to find a way to replicate the smell of said find for the dog's toy. You shouldn't pack valuable artefacts in a chew toy after all.
Should I ever find a job as a professional archaeologist (for, contrary to Martin's flattering introduction I'm a mere student of archaeology) I might still try that someday.

And of course it's always more fun to have a dog around at a dig.

Oh, this is fun. It could work actually, it boils down to how distinct the metal odor is (e.g. how is a metal object separated from metals present in nature--let's say, iron that is present in rocks as opposed to iron in old man-made objects?--even thought the smell of human handling of the object would remain a while, that's sure no help in this case... It's quite different when the police and rescue teams use dogs to find relatively recently lost/dumped objects). I would think blindfolding is absolutely unnecessary (it sounds really like something from the 16th century ;)), and also I think modern dog trainers could use the same sort of training equipment that is used to train dogs to distinguish between smells or detect smells. (For those who know Swedish, one name for it is "preparatsök" and you can take courses in it at Hundcampus, a facility in Värmland, which has special training appliances, so you really can keep the training objects/substances untainted by dog-owner smells etc!)

Smelling jobs are what dogs excel at, fortunately!

Do metallic finds even have a smell once they've been buried long enough to stabilise?

This makes me think of the workers who were supposedly employed in the mints in Imperial Rome to test the quality of coins by smell. That sounds crazy, but one of my colleagues who is spookily good at metal diagnosis, and who told me the factoid about the mint workers, certainly does smell the coins in question, and there is a definite tang to copper versus silver that even I can pick up, unless the pieces are cold. So I don't see any basic problem with the idea of taking advantage of a far better sense of smell. What I don't see is the advantage over metal detectors...