He\’s going to get panned for telling the truth

Defending supermarkets as a whole, he added: “OK, you can say we haven’t been testing for horse – well, why would we?

“We don’t test for hedgehog either.”

That is a very interesting point actually.

This is analagous to testing for metals. For some metals you can do a simple test and just go \”Yeah, that\’s 99% iron\” or whatever. Look, feel, shape, density, magnetism, a skilled scrap yeard worker can tell what most of the stuff coming through the yard is in seconds. Add a Geiger counter for radioactivity and you\’re pretty much good to go.

But then there\’s another end of the industry. Where you want and need to know not just what is the material basically, but what are all the trace elements in it? At which point you start to test for 72 elements (you can miss out the noble gasses and the transuranics etc). For example, we have one product where we need to make sure that Fe is below 5 ppm, Zr below 1 ppm. It\’s a 99.995% purity and we don\’t much care what the other 44 ppm of contaminants are. 999,950 parts Sc and O, 50 ppm allowable contaminants, but that Fe below 5, Zr below 1.

Given the way such measurements work you don\’t actually count the Sc and the O. You count everything else and subtract from 1 million.

As you can imagine, this testing is rather more expensive that the glance and a feel that happens in the scrappie\’s yard.

Now let\’s turn to that meat problem. We\’re going to test something to make sure that it is indeed what it says. Most of the time, usually, we\’d go looking for beef DNA and on finding it say, yup, that\’s beef.

But now we\’re talking about trace amounts of other species. Some of this horse contamination is someone deliberately substituting, yes. But a lot of it, those trace amounts, is someone not cleaning the pipes between species being processed. Or the knives even. Which leads us to something of a problem.

How many species do we test for? Some minced beef…..or pink slime perhaps. Do we test for beef and horse? For beef, horse, mutton, pork, chicken, duck, goose? What about rat and mouse? For I\’ll guarantee you that however much people try there will often be the odd molecule of either one of those in there. Sparrow? That\’s more of a problem with grain processing but still.

For example, one lovely story about vegetarianism. Those (umm, OK, some) who have moved from the sub-continent to the UK. They carry on eating the (possibly Hindu caste based) vegetarian diet they are used to. And they start falling prey to all sorts of dietary deficiencies. Anaemia, there have even been reports of kwashikor (a protein deficiency). The grains and the pulses of the sub-continent have rather more insect and other residue in them than our more modern processing and storage systems provide.

People don\’t test for hedgehog DNA in meat supplies, no. But how many species should they test for?

And there\’s one other thing. Given the accuracy of today\’s tests what are we actually meaning when we say \”trace\” amounts? Some modern water tests for example go down to parts per billion.

And one thing that I am absolutely certain about is that if we start testing food supplies down to ppb then absolutely everything is contaminated with something. At ppb you would, for example, find human DNA from the skin flakes of the human workers doing the processing. Does this make us all cannibals?

16 comments on “He\’s going to get panned for telling the truth

  1. Yes, this problem has already arisen in food supplies. A hyper sensitive test (ppb) for artificial dyes was discovered. Suddenly a lot of (probably all) chilli powder was found to have artificial dyes in it. One batch was fingered.

    There had previously been some genuine adulteration going on, but this was so microscopic that one theory was that it was traces of dye from the labels on the sacks.

    Anyway, the FSA ordered a large recall of food containing any chilli from the “contaminated” batch. It then had to back down and say that people could go back to the old test, as pretty much all the processed food in the country contains some chilli/paprika for taste or colour, and it was all “contaminated”.

    Which led to the strange position that the same stuff would be uncontaminated if you supplied it before the new test came in, then contaminated if you supplied it after the new test came in, then uncontaminated again a few weeks later if you supplied it after the FSA said you didn’t need to use the new test.

    Good for lawyers, not so good for anyone else – google Lion Foods v Hazelwood Grocery

  2. I suppose a standard where you specify maximum levels as you might use in metals might make people a bit squeash. Titanium with no more than 0.2 wt% oxygen is one thing but beef with no more than X % human and no more than X % rat might be a bit off putting

  3. beef with no more than X % human and no more than X % rat might be a bit off putting

    in marketing terms, you’re right of course. But that’s because people don’t understand anything beyond the basic “that’s yucky!” Point out to people that when they can smell a fart that’s actual turd molecules going up their noses and they have a really quite visceral reaction* but break it down into the names of the compounds and suddenly it’s fine again.

    The air that we breathe presumably contains trace elements of rat and human, after all.

    *have more fun: remind them that if they flush the lavatory without closing the lid then all those molecules get liberally dispersed about the bathroom, and may well end up on their toothbrush…

  4. The air that we breathe presumably contains trace elements of rat and human, after all.

    Hence the whole “how many molecules from Julius Caesar’s dying breath do you breath in …” question. Which, of course, relies on Avogadro’s number being quite as enormous as it is (and Gaius J have lived so long ago you can make assumptions about mixing.)

  5. “beef with no more than X % human and no more than X % rat might be a bit off putting”

    I see your point, but there are already similar rules about the amount of arsenic you can have…

  6. Pingback: Opening the food testing can of worms: “We don’t test for hedgehog either” « Quotulatiousness

  7. JuliaM,

    And why are we still talking about this and not about the (fatal) failings of Mid Staffs NHS staff?

    Because the government don’t want to.

    I’m hoping that it’s just a case of Hunt keeping his powder dry. Let the lefties come out and defend the NHS, then hit them later with the report into the 12 other trusts under investigation (which you can bet are as bad). My guess, it’s going to be the moment when the quasi-religious love of the NHS evaporates.

  8. Presumably a supermarket would just test for beef on a beef product and if it was less than 95% or whatever, then they’d reject it regardless of what species the remaining 5% was, horse, hedgehog or quagga.

    I worked on quality in a sandwich factory and once rejected half a ton of beef slices for being the wrong shade of grey, you don’t really need much of a excuse to reject something that could be dodgy.

  9. There is no point testing beef for beef. There is bound to be some. But if you want to actually catch people – and I do not believe that is the purpose of modern regulation enforcement – then you ought to test for the obvious.

    We have a lot of horses and donkeys in Britain and Europe. Many of which end up in slaughter houses. And which look a lot like beef.

    I would have thought a horse meat test every now and then was obvious.

    Some metals are tested. TW has written here before about Zirconium being too pure for instance. Steel is routinely tested for radiation before being recycled. There is no point testing for everything, but for some things, testing is sensible.

  10. Is that really how it works?

    I can understand that if you want to test a metal for multiple elements you are going to need multiple tests, but different types of DNA are still DNA. I expected them to take a number of samples from a batch of meat and test each of these samples against a database of reference DNAs. The database would contain cattle, chicken, etc, plus any other animal you wanted to test for.

    Admittedly my understanding is based on CSI, but I’m pretty sure when a real-world Hodges tests a DNA sample against their crime database he doesn’t perform 3 million individual tests.

  11. Well, whatever that spokesman said somebody somewhere obviously was testing for horse. Otherwise they wouldn’t have found it.

  12. “well, why would we?”

    How about, “because you source your beef from a long and complicated chain which also has horse meat (but not hedgehog) flowing through it, you fucking cretin?”

  13. @13: now that’s not how it works. This isn’t CSI. You can take a sample and test it, say, for human DNA. But you can’t take a sample and say, “what is this?” It can be tested against all the things you might be looking for, but if there’s DNA there from something you’re not looking for, it doesn’t show up. It’s not like you put it in a magic DNA fingerprinting machine and it spits out a pie chart with x% DNA type 1, y% DNA type 2 etc..

Leave a Reply

Name and email are required. Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>