Measuring Things

In blog discussions recently I’ve heard it said that most things can’t be measured. Most interesting things, anyway. Things like beauty and love and morality. This is slightly puzzling. I think it is probably a failure of imagination. I’d like to explore measurement a little.

Why Measure

Measuring things sounds, to a certain audience, like a horrible thing to even want to do. If I measure the beauty of a piece of art, then I have a number, rather than the piece of art.

This, frankly, is a silly point of view.

Measurements are another form of description. If I could say that a piece of art had a beauty score of 9.4, then this would no more summarise that piece of art than me saying it was 20 inches wide, or an oil painting, or worth $15m. Any kind of measurement is a particular feature, one of many. The original piece of art is still the original piece of art, no simple summary or list of predicates is going to replace it.

But measurements are useful for understanding things, because they produce quantities, and quantitative reasoning is more powerful than qualitative reasoning. We have far more tools to do that work.

So I think we have to ignore the straw man that says things like “you can’t reduce Proust to a number”. Of course not, any more than you can ‘reduce’ any phenomenon to its description. You can, however, measure Proust in various ways. And those measurements can tell you something new and interesting about his writing.

How to Measure

To measure something, you need to be able to say what you mean. Let’s take beauty, for example. Well there is a whole philosophy of art, but you don’t need to resolve any of those questions to proceed: you just need to choose some understanding of what beauty is.

So, for example, you might say that beauty is a subjective response to a piece of art, made by people who understand the work. You might go on to say that such understanding takes some degree of expertise. So you’d pay more attention to the response of someone who speaks english when considering the beauty of a Shakespeare sonnet. And even more weight to someone who reads a lot of poetry.

Well subjective response can be measured, in very many ways. We do this as a society all the time, at all levels. Elections reduce the subjective response of a population to a number of votes, surveys, customer research, focus groups, AB split testing: we have very honed tools for doing this.

How about understanding or experience? Well we measure this all the time too. It isn’t hard.

There is nothing difficult that I can see in measuring the subjective response of individuals with understanding towards a piece of art. There is nothing difficult, therefore, in measuring its beauty, under that understanding of beauty.

If you have a different understanding of beauty, follow it through the same way. If you can’t decide between ten different concepts of beauty, well combine the ten different measurements (there are tried and tested tools for doing that combination too). If you think beauty is actually a combination of several different underlying criteria, well measure each criteria and give a multi-dimensional result. Or do a PCA and find what the underlying variables in beauty are.

You may not want or need to do this measurement – but that doesn’t mean it is not possible.

I struggle to see why so many people reject empirical knowledge out of hand. Its as if it has become so associated with geeks in lab-coats that it creates an irrational fear of looking stupid. Which in turn (as all such fears do) engenders hostility.

Advertisements

12 Comments

Filed under Uncategorized

12 responses to “Measuring Things

  1. atimetorend

    “You may not want or need to do this measurement – but that doesn’t mean it is not possible.”

    I agree, possible, so the statement you are critiquing is wrong, but perhaps it would be more helpful to just say that measuring some things is possible, but relatively trivial in the human scale of things. But I bet that is a normal part of everyday life. For example I might spend more time looking at a picture in a book which I find more beautiful. I may have quantified beauty on an emotional or subconscious level? Or how did I choose my wife? So maybe not so trivial?

    I struggle to see why so many people reject empirical knowledge out of hand. Its as if it has become so associated with geeks in lab-coats that it creates an irrational fear of looking stupid. Which in turn (as all such fears do) engenders hostility.

    Could rejection of empirical knowledge out of hand be a reactionary response to an overvaluing of empirical evidence? Your description of the phenomena sounds reasonable, but like anything people will take that and run too far with it, having more confidence than is warranted in their empirical claims. Which leads to bad interpretation of the data at best, or fundamentalism at the worst.

  2. Ian

    @attr

    Hmmm, I agree with you, I think.

    I think we have to differentiate between descriptions and phenomena. A measurement is just one description of a phenomenon. It is a type of description that provides us with useful tools, but it isn’t otherwise special. Would someone say it is trivial to describe a piece of art? I guess not.

    If not, then what is special about measurement that makes it particularly problematic? It knowing that people have double the emotional reaction to piece of art A, vs. piece of art B really a less useful piece of knowledge than that the piece was created using oil paints?

    And as for using empirical data wrongly, or overstating it. I agree. There’s nothing about measurements that guarantees your reasoning is going to be sound and your conclusions good. I wanted to make the point that it wasn’t the use that was being criticised. It was this notion of measurement in general.

    I wonder if part of the resistance is something to do with being scared that the numbers will contradict the story we tell ourself. If the numbers show I love my wife less than you love yours, for example. Dunno.

  3. atimetorend

    Maybe it is not that something *can’t* be measured, but maybe people are afraid that measurement itself will trivialize something they feel to be very important. If you use numbers to show how much you love your wife you are in danger of trivializing the meaning of your love for her. At least you would have to be *very* careful in expressing that to her (or I would to my wife anyway).

    Ever read this book?

    …[Little Nutbrown Hare] was almost too sleepy to think any more. Then he looked beyond the thorn bushes, out into the big dark night. Nothing could be further than the sky. “I love you right up to the moon,” he said, and closed his eyes.

    “Oh that’s far,” said Big Nutbrown Hare. “That is very, very far.”

    Big Nutbrown Hare settled Little Nutbrown Hare into his bed of leaves. He leaned over and kissed him good night. Then he lay down close by and whispered with a smile, “I love you right up to the moon – and back.”

  4. Boz

    Emotions can in principle be measured by monitoring the activation of neurons in the brain.

  5. Ian

    Yes, it was a favorite with our son!

    I agree, if I reduce my love for my wife to a number, I would trivialize it. But that doesn’t mean it can’t be described by (one or more) numbers. I wouldn’t trivialize it by describing it as, say ‘tender’ would I, then why would saying seeing her increased my arterial blood flow by 25%? Hmm.

    I guess I’m really arguing with myself. I do see the difference, but I’m struggling to understand it.

  6. Ian

    @Boz – Yes, exactly. And hormonal responses, and physiological correlates, and so on.

    Welcome to the blog, by the way!

  7. I totally agree, Ian. But you probably know that.
    Now, is this position given a proper name in Philosophy? Is there an opposite school?
    My wife will avoid arguments because she knows she will sound dumb to herself when trying to think on her feet. She does not like that feeling. Perhaps some people fear measurements the same way. They fear others will use them to imply something that threatens them and they won’t be able to argue against them. After all, real good use of numbers and measurements takes training and discipline and some native intelligence. Not everyone has those. A certainly, as ATTR says, numbers and measurements are used to manipulate people all the time.

    Let’s say someone who is real good at math and manipulations puts a whole numerology thing together to convince someone, numerically, that the world, by Mayan calculations will end in 2012. Now, a somewhat paranoid superstitious person who is weak in math on hearing it decides to cover his ears saying “Nah, nah, nah, nah, nah, I ain’t listening, you can’t measure everything.”

    You see, he is afraid if he listens he will be stuck in very uncomfortable cognitive dissonance, he will probably fall for the numerology but he doesn’t want to start saving up supplies and quit his nice job to join a spaceship to take him away.

    Or imagine someone who was raised by loving Christians, married to a beautiful loyal Christian wife and starts to be shown measurement ways to show false dating in the bible or authorship of his favorite texts. He covers his ears. He knows he does not want to be thrown into cognitive dissonance.

    That is one of my bets on why “they” do it.

  8. Im sure at some level everything can be measured. I would imagine for some people this is very important and for others not so much. This discussion reminds me of a friend of mine and how we play pool(billiards). He is very methodical, he calculates his every shot. The angle, the spin of the ball, right down to the millimeter. If you were to ask him how he does it he could explain it quite succinctly. He is an excellent player and we have great matches. If you were to ask me how I play my game the response is quite different. I would say “I just feel it.” 🙂

  9. Ian

    Yeah, sabio and t4t. Lots and lots to think about.

    I’ve seen that very situation play out with climate change, Sabio. Someone who didn’t want to know about the science, because they (rightly) said that the people on both sides of the argument were wielding scientific arguments beyond their understanding and knowledge, therefore they had no hope of deciding which was right.

    T4t – The same thing plays out with language, I think. I know a little bit of a handful of languages. But it is only English that doesn’t take a conscious effort. I know some people learn languages by ‘feeling’ it, repetition particularly. I can’t do that. I have to analyse, explicitly learn the rules, and apply them consciously. I suspect that means that I’m much faster at getting to a very basic level with a new language, but much much slower at becoming fluent (which seems to fit, since I’ve never managed to become fluent at any other language, despite trying).

  10. atimetorend

    Not only can you measure ANYTHING, but it turns out you can formularize it as well:
    http://www.wired.com/magazine/2010/01/pl_arts_found/

  11. I totally get this! What I was trying to say in the argument was not to rely totally on this… nor should we totally rely on the other. you strike me a someone who would take an integrated approach..so would most on your blog and sabio’s. i’m unsure about boz and a few others… but the scientific method is a great one, prolly one of humanity’s best ideas.

    love the name of the blog btw. fantastic!

  12. Ian

    @attr – excellent – I love the sand pile particularly.

    @Luke – cool. Yes I think integrative approaches have to be pursued. The title says it all 🙂 Its been a good comment thread this, moved my thinking along a bit, I think.

    Great to see you here! I was hoping you’d pop in now and then 🙂

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s