Universals Crash and Burn

As we observed in the last post, John Locke attempted to update Aristotle's view of universals by saying that we select key attributes for defining categories based upon our objectives. So, we can look at games, decide what salient features they all have in common and define the concept of game based upon those features. It looks like the project might be gaining altitude again. But, instead it will take a nosedive.

David Hume would observe that we form categories based on a subconscious cognitive process that cannot be made explicit. That is, we go about our business in the world, experience games, and form a visceral concept of what a game is. Any attempt to explicitly define characteristics would be artificial and doomed to failure. The Supreme court justice who admitted that he could not define pornography but claimed - "I know it when I see it", was embodying Hume's view of universals.

If Hume steered the project into a nose dive, Wittgenstein let it crash and burn. According to him, the only thing the elements of some categories have in common is that they are members of the same category. He used a, now famous, analogy to family resemblances. If you go to a family reunion and look at the members of the family you will see common elements. Some will have the family eyes. Some will have the family nose or mouth. Some the family brow. And so on. There are overlapping features held by subsets of the family but no set of features common to all. For example, not everyone has the nose. Or not everyone has eyes. Universals, according to Wittgenstein, are made the same way. You cannot define a core set of attributes because there is no core set of attributes that all instances have in common. And the concept he used to illustrate this was - you guessed it - games!

Should we despair at this point and give up our attempt to define what constitutes a game? Should we despair that if this is true knowledge is not really possible? No, that would be a little overly dramatic. Instead we will back up and see where these great minds went wrong. We will go back to some road signs provided by Bacon and Aristotle and try to get back on the right path. And that will be the topic of the next post.

The Decline of Universals

Aristotle saw universals as constructs based in reality. A tree is a tree because it shares key attributes with other trees. Anybody can look at a tree and a squirrel and see that they are not the same thing. Initially this seems to solve the problem. But in the long run it only makes it worse. It answers the question - where do universals come from? - by saying that they are constructed based upon common attributes. However, it then leads to the question - where do attributes come from? If the squirrel happened to be brown, we could create a universal "brown things" to which the tree and the squirrel would belong. Aristotle might balk at this by citing that "brown" is not an essential attribute and universals should be formed based on common essential attributes. This sounds good and it seems like we are making progress. But, we are not. How do we know that an attribute is essential? Is an attribute essential because it is in the definition of kind? Or is an attribute in the definition of kind because it is essential? That whirring sound in the background is our cognitive wheels spinning in the philosophical mud.

Locke made some progress by saying that the creation of universals involves a cognitive process of abstraction whereby we we select attributes based on some set of objectives. I am going to jump in and help Locke out here by saying that our objective is usually intellectual economy. We create categories because they are efficient ways to organize our knowledge. Otherwise, we get into huge trouble with Locke. We get into trouble because it raises questions such as 1) how do we select objectives?, and 2) how do we know that a particular category is better than another category at meeting those objectives?

Locke has both helped and hindered the pursuit of universals. He has helped in that his claim is intuitively appealing. It does seem to be the case that we look at a collection of particulars and though some cognitive process extract a group based on similarities. In fact, it feels so built in that one has to wonder if it is just a function of the brain. He has hurt in that it has made our understanding even more unclear. How does this process work? Can it be made explicit? How are objectives defined? How are categories evaluated with respect to objectives?

All we were trying to do was to figure out what a game is? How did life get to be so complicated. Any six year old can tell you whether or not something is a game. And, yet, a philosophical adult has to admit defeat. Why is that? Well, David Hume will come to the rescue on this question and, in doing so, take a stab at the larger question. However, as we will see, Hume will jump into the muddied waters and stir up even more mud. Following that, Wittgenstein will show that once you are into muddied waters you can't get out. And that is what we will take up next.