Results of our experiment on soup categorization

On this page, we will discuss some hopefully interesting observations about the data gathered from the players of SOMETHING SOMETHING SOUP SOMETHING. The most obvious among them is that players had (sometimes wildly) different conceptions of what soup is. As mentioned in the '... but why?' section of the website, this resonates with our intention of revealing through gameplay "that even an ordinary concept like ‘soup’ is vague, shifting, impossible to define exhaustively."

A less obvious observation has to do with a difference that we managed to identify between the conceptions of soup of our initial focus groups and those of the players. To illustrate this point, let us look at the formal property of soup that was the most cited in our focus groups: its comestibility.

• Of 21 participants in our focus group, only 1 did not mention comestibility in any way as an essential component of ‘soupness’.

• Out of 29 data points obtained from the game in an experiment setting, 5 participants defined soups as being necessarily composed of only edible ingredients. 7 data points in our players’ group accepted dishes that had inedible components that did not change the soups comestibility, such as cocktail umbrellas. The remaining 17 data points out of the 29 players tracked during our experiment allowed for ingredients that signified contamination or inedibility, such as leaky batteries or stones.

We recognized the same pattern when considering the second most often cited formal property of ‘soup’ according to our focus groups: its being liquid or liquid-like. 17 of our 21 focus group participants indicated it as a formal property of soup. However, 13 of the 29 tracked players accepted solids (such as rocks or ice cubes) as viable bases for a soup. The results still show that liquids are perceived to be ‘soupier’ than solids (all players accepted at least one liquid as a soup). However, having just under 45% of players accepting solid soups shows once again how soup definitions neglected items that are not immediately obvious as soups (i.e. lower membership items).

This phenomenon is even more pronounced when we observe formal properties of soup that are less frequently mentioned by our focus group. The third and fourth most common characteristics were respectively that a soup is to be found in a bowl / bowl-like object and is eaten with a spoon. Almost 25% of the focus group respondents mentioned these properties. Looking at players’ data, instead, these formal properties appear to be hardly - if ever - considered. Just over 10% of our players required soup to be served in a bowl-like container (a bowl or a cup), while none of them specified the need for it to be specifically served in a bowl. Only 1 player required soup to be eaten with a spoon-like object (spoon or ice-cream scooper), while none of specified the need for it to be exclusively eaten with a spoon.

Definitions of 'soup' are established by what is most commonly present in (or with) them. We noticed that, when presented with different soup examples, people were more conceptually inclusive than they were in their defintions.