# Can we really measure something that's subjective?

Someone said we can't measure how creative something is, but isn't that completely wrong and dishonest?

Isn't it the same as for intelligence? Isn't measuring intelligence based on a subjective model of intelligence? For example, you need to divide up intelligence into components subjectively and then measure these parts to come with a measure of intelligence such as IQ.

I only asked if there was a way to measure how much more creative a work by Chopin would be compared to a work made by Bach using computational creativity, and someone disingenuously said it was impossible, because it was subjective, but we just need to make a model that allows us to measure how difficult it would be for a computer to generate the same music, and divide it into different components?

Is there anything wrong with what I said?

The concept of 'measurement' depends on a collective agreement on a 'metric'. In other words (to use Wittgenstein's example), to say that something is one meter long we must all agree that some given thing is (circularly and arbitrarily) defined as a meter. That was originally a small fraction of the circumference of the earth, then a 'standard meter bar', now the distance covered by a certain number of wavelengths of an emission line of krypton-86. Note that each time the metric changed, the very definition of a meter changed, so that everything that used to be precisely a meter long no longer was.

With that in mind, the difference between 'subjective' and 'objective' is follows:

• 'Objective' means that we have agreed to some (arbitrary) metric
• 'Subjective' means that we have not agreed to some (arbitrary) metric

We call creativity 'subjective' because no one has created a (formal) standard we might use to measure people's performance on 'creative' tasks. If someone were to invent such a metric for creativity, and then convince everyone to agree to it, then creativity would become 'objective'.

You can create a model that measures how difficult it is for a computer to generate music, sure. But then you have to offer that model up as a standard and convince people that it is a good and useful standard. That's much harder, because you have to argue successfully that what your model measures is actually related to what we commonly consider 'creativity'. If you can get people to (circularly) define creativity in terms of your model, then your model will become an objective measure.

Crazy...

As Ted Wrigley's excellent answer notes, measurement require consensus on definitions, in the sense of rendering "finite." IQ tests "measure" something, for example, but is it really "intelligence"? Yes, if that's how we all want to limit the definition of intelligence, which in fact we generally don't!

In a sense, you can "measure" anything, turning qualities into quantities, but nothing with absolute precision, as quantum conundrums show, or convincingly. For a creative or "subjective" being to "measure" subjectivity means to apply some finite metric "from the outside," so to speak, and suggests an infinite regress. Once the subject has completed "measuring" itself, it has already changed.

One can, of course, measure "creative" output in the sheer quantity of pages or notes or whatever. (And sheer quantity is usually an aspect of what we call genius--Dickens, Bach, Picasso, etc.) But everyone agrees we are really talking about the "quality" of the production as well, or a Xerox copier is greater genius than Dante. And here the ambiguities flourish.

Indeed, there may be something contradictory about the whole idea of "measuring" creative output, a kind of Heisenberg puzzle. The more precisely you measure "creativity" the less "meaningful" it is. As Luhmann defines it "meaning" is a kind of relation of possibility to actuality. As an artist, say, "actualizes" possibilities one can objectify the work, but one cannot objectify or quantify the further "possibilities" in relation to the actualized work.

Have to admit, I'm not entirely satisfied with this answer, but pencils down. Good books on such topics include Postman's "Technopoly" and Gould's "Mismeasurement of Man." Not to mention the related and extensive debates on "qualia"