The Broken Brooch Fallacy

           Current arable intensity (Nature)

James Slater, detectorist, in a  comment to an earlier post here seems to be under the impression that he's telling us something we do not know, but in fact has not first himself checked out his 'facts'. I'd say that was a pretty unintelligent thing to do (especially when trying to deal with the arguments of a guy it says over to the left in the sidebar has been studying artefact hunting and related issues since the 1970s).  Among the mantras he's heard others chanting and decided to parrot here on my blog, we find this one: 

Many artifacts (sic) such as Roman brooches come out of the soil smashed to pieces by power harrows and other farm implements. If we have an opportunity to recover thousands of long lost artifacts (sic) and coins and save them from certain destruction then we should take it [...].

Well, this is an object-centred argument because those brooches are not in fields in a vacuum, and this blog is about archaeological evidence being trashed. 

But above all, I'd like to see that 'statement of fact' documented. This is quite important, because this argument is so frequently trotted out by those who've only got anecdotal evidence at best to support it. Now, I am aware of papers where it has been shown that pins (hairpins, clothes fasteners), long, thin inherently fragile, especially when corroded, get broken in the ploughsoil as opposed to excavated examples (I'm not going to go into it here, but in one of the most frequently quoted there's a rather awkward methodological flaw). But the tekkie (and tekkie-supporting archaeologists) argument is that "all" artefacts suffer this (so the mantra is "better out than in"). Roman brooches are an ideal test case, they've often got a long thinnish 'body', with at one end a protruding catchplate and at the other a strongly protruding headplate with the spring and all the other gubbins to create friction in movement.  Ideal for testing what happens when a plough moves them around. I'm glad Mr Slater mentioned it.

Let Mr Slater take all the Roman brooches sold by British sellers on eBay (the ones artefact hunters don't want for their own collections - and ignoring the many fakes there) and all the Roman brooches on the PAS and UKDFD databases and (if he likes) all the ones displayed today on the metal detecting forums and facebook pages, and lets see the figures. Together there will be a sizeable sample of several tens of thousands. 

A cursory look shows that most are relatively undamaged, but since Mr Slater insists, we'd like to see figures for: 

Complete or near complete (ignoring the pin, these usually break anyway) n=??

Fractured - ancient fractures n =??, and  

fractured, fresh breaks that do not affect just the pin n=??. 

We would like to see some numbers to back up this statement, please. Then we can talk about facts and not mantras. 

Since I very seriously doubt that any metal detectorist will ever produce an argument they can back up with actual evidence, instead of anecdote or appealing to 'common sense', I'll give readers a clue. Roman brooches on the PAS database , total quantity: 32,568. BUT  the ones that mention "plough" in the full description made by the FLOs is... just 528.

What that means is only one in 62 is in some way described by the PAS (who also use this same argument as Mr Slater) as affected by the plough. 

But then look at the actual fibulae in that group, they are not broken, many of the descriptions claim they are somehow "eroded" by the plough. And looking more closely at the photos of a selection of those objects, one is prompted to ask just how many hours training in soil and corrosion chemistry FLOs get before they start work... 

Now look at the map of them. There is something wrong here... obviously. There are a couple of FLOs that say in their descriptions they see this effect,* and about thirty more that don't. Yet those that don't are those that actually operate in the areas of England and Wales with the most intensive ploughing (and these patterns are probably going to change soon anyway, when broken brooches will be the least of the problems). 

For the record, the total for "agricultural machinery" is... 2. (and just for good measure "artificial fertiliser" another common argument - zero).

So, Mr Slater, and all those that want to parrot the same mantra, either produce some proper figures, or please stop trying to use this false argument to justify artefact hunting. 

* It's worth noting this. Just taking the top five (they tend to cluster chronologically - sadly the records are now anonymised so we cannot see which archaeologists are noting this effect): 

Staffordshire (132)

Shropshire (129)

Worcestershire (79)

Warwickshire (62)

County of Herefordshire (48)