Attempt 1

What is soup? No seriously, try come up with some ideas of what makes soup soup; you probably were thinking some of these:

• Liquid
• Thing you put in a bowl and can consume with a spoon
• has flavourings dissolved into the liquid
• made of natural ingredients

The problem is if you’ve said those or similar… A cup of seawater is soup. I can put it in a bowl, it’s got minerals dissolved in it and it’s full of organic living things like plankton. It’s not tasty but I definitely could have a spoonful. Let’s try again to define soup, but making sure my cup of seawater isn’t soup, because I don’t want to eat it.

Attempt 2

New soup rules:

• tasty
• hot
• contains vegetables

The problem is we’ve failed again. Gazpacho doesn’t count as soup anymore because it’s cold. If I make my soup badly and put too much salt in and it tastes vile it’s suddenly not soup anymore, and a chicken noodle soup made from chicken stock is also not a soup anymore.

Who Cares

Clearly these things are soups and clearly ocean water isn’t a soup, but coming up with rules to separate the two exactly is extremely difficult and this is not just a problem about soup, in linguistics, philosophy and science alike it’s a real issue trying to precisely define categories even when they are trivial intuitively.

There is a bigger question here though, who does this actually matter for? Despite my failure to neatly and wholly define soup, we all get what soup is and if someone invites you round for soup you can be pretty sure it’ll be carrot or tomato rather than seawater. For matters greater than soup though this problem does matter. If you’re describing a scientific result in a research paper, the ability to replicate your results is essential and that requires a precise clear communication of what’s included in a given group and what’s not.

In my upcoming long piece on “The Species Problem” I will delve further into a practical example of this definition difficulty. Often the practical solution used in disciplines where precise definitions do matter, is to stray into wordier longer descriptions. This is how “Cows” becomes “Domestic Bovine”. That matters when describing say, a disease outbreak study, because while the groups “cow” and “domestic bovine” have a very big overlap, there is some fringe difference. The latter would include farmed Bison in the USA or Water Buffalo in India. This, of course, does have its downsides; but often is the best solution we’ve got to this ‘Soup Problem’. In another upcoming piece I will discuss more how this shift in terminology that gains precision, costs us readability and the accessibility of scientific research to the masses.