Friday, March 11, 2011

Book Review - The Structure of Scientific Revolutions by Thomas S. Kuhn

Disclaimer: This is rather a mix between a review and a discussion inspired by this book.
Thomas S. Kuhns book "The Structure of Scientific Revolutions" proved to be one of the most boring toughest books I've read in quite a while. Kuhn aims to explain how scientific revolutions come to exist, what they change and how they work. Although the subject is interesting, the language of this book is extremely dry and technical, and reading full sentences without losing track midway proved to be the real challenge. You might ask why I even bothered with this book in the first place?

I was reading another book, "Black Holes and Time Warps" by Kip S. Thorne, a great book that I strongly recommend by the way, in which the author in passing by mentions Kuhns book as one of the most "eye-opening book he's ever read". That got me interested of course. If this great author thinks this book is good, it has to be good! The title of the book also made me interested. I've always been fascinated about how science, which should be completely untouched by human flaws, is affected by human nature. Alot more than we think it is of course.

Kuhns book is alot like the candy bag of a friend. There is alot in there which will make you frown, because your friend doesn't have the same taste as you, but here and there you will find some tasty bits. This is not to say that I agree or disagree with what Kuhn is saying, but that the interesting stuff is covered in alot of heavy text and I could think to myself that I should just give it up, plenty of times. And just when I did, I found something really interesting and had to continue reading (the book also was a great way to help me fall asleep in the evenings). Not quite unlike doing something boring in WoW. Just when you think that it isn't worth the time anymore, something interesting happens and you decide to give it another half an hour of your time. Also the book is rather short!

If you have the patience, Kuhns does a good job in explaining what exactly it is that pushes the scientific community from one "reality" into another. From accepting one thing as true to suddenly accepting something else as the new truth. How do we distinguish between these truths? When do we decide that the old truth doesn't work anymore? One could imagine that people would abandon the old ways as soon as they prove to no longer accord with observable reality, but this is not so. It often takes several generations from the first discoveries until the final step towards a new paradigm is taken. Why is this so?

One of the reasons Kuhn mentions, and which caught my interest the most, is how we interpret reality. I read this book in swedish so these quotes will be amateur translating from me.
Kuhn talks about how we can't really say that there is only one definite reality. Even if there were (which really is a philosophical question) there is no way to tell whether we experience it as it really is, without interpretation, or not.

"Is sensorical experience fixed and neutral? Are theories simply an interpretation of given data?" -Kuhn p. 106

We often assume that person X will interpret a set of data in the exact same way as person Y. The numbers are the same, so the conclusion must also be, right? This is in fact rarely the case. Not only do we interpret things differently, we even correct things in our minds unconsciously. We might think we see what we see, but we don't, in the sense that we can all agree upon it. Kuhn gives an example of a psychological test, where the test subjects were shown various playing cards. Most of the cards were normal cards, but some of them were special. They could be black hearts or red spades for example, cards that don't usually exist and that most people will never have seen. They showed the different cards, both normal and special ones, to the test subjects for only short glimpses. All the test subjects said they had seen normal cards. So they increased the time frame in which the test subjects got to see the cards, but everyone still said they only saw normal cards - even when they didn't Not until the time frame was extended to several seconds did people start to become uncertain. They still didn't want to believe that they saw anything but normal cards, but said that something looked "weird" about them. But they still didn't notice that the cards were in fact not what they thought they were. When the time frame had extended towards 10ish seconds, people started noticing that they were in fact not looking at only normal cards at all.

We learn how to interpret reality, we fit things into categories to better understand how the world works. We have to do this, because if we had to re-categorize everything we saw each time we saw it, we'd work very, very slowly. We'd be like a desktop filled with documents, instead of a few folders filled with documents. Even if it means a couple of documents end up in folders that aren't completely right, it still is a lot more efficient way to handle input. In our daily life we have to assume that what we think is true, really is true. We can't go around questioning everything we think about reality. This is one of the reasons science work with "new truths" each day. We need a foundation to stand on, but we must also be ready to dismiss that foundation if that would suit us better.

This goes for science as well. Exploration and discoveries would be extremely slow if we had to re-affirm the truth about every step we take on the way. And as long as the measurements are crude, we will see one reality. Not until the measurement become more exact (longer time frames) will we start to consider that we're actually looking at something else than we thought we did.

Kuhn also talks about what could be described as "trends" within science. Even if someone does manage to make a discovery that goes against the old "truths", it doesn't always hit through with the rest of the community. Like I mentioned, it can often take several generations before people have decided to take the next step. Sometimes science just has to be ready for the change for it to happen. One example (not from the book) is the idea of evolution. Although Gregor Mendel had written an essay on hereditary traits already in the mid 19th century, it wasn't until some 40 years later that people actually started caring about it. Science was ready for genetics then, but not before. There are tons of examples like this one where you can see how "discoveries" actually can be found several hundreds years before as theories by various scientists, that during that time simply went unnoticed. Who decided which science is worth putting time and effort into?

"Science doesn't busy itself with every imaginable laboratory experiment there is. Instead some are chosen as more relevant for the unfication between a paradigm and the immediate experiences that this paradigm partially has chosen". - Kuhn p.107.

Science chooses all the time between which science is important right now and which isn't. In the early days, when alot of scientists had maecenas or benefactors who financed their studies, their choices were partially being governed by the interests of those benefactors. This is still true, or maybe to an even bigger extent, today. The "important" science is often that kind of science in which someone (usually companies) can earn money from. The science which doesn't earn money, so called "ground science" is usually financed by the government who has less financial interest. This science often includes basic mathematics and space exploration for example. Science in which people can make money is just about anything that can be converted into medicinal or technical (like machines) use. And eventhough money definitely is a big factor in which science is being chosen as important, it's not the only one. The science community will have different subjects which it holds as more important and interesting, seemingly arbitrarily. Yet, science has to focus on something of course. Everything can't be interesting, because then nothing would get enough attention to actually get us somewhere.

Overall I don't recommend reading this book unless you're really into this subject. It is interesting, but there are plenty of other books that present these kind of issues in a much more enjoyable way. Nothing wrong with the content or the lines of argumentation, but the language is just dreadfully boring.


  1. If you're into the topic of communication and psycholinguistics, taken from what you wrote about how we interpret reality and messages, I can really recommend Paul Watzlawick's axiom theories and F.S. von Thun's 4-side model. both are widely acknowledged and I've personally found them to be the most insightful and apt when it comes to understand and judge human communication.
    Schulz von Thun isn't nearly as dry as Watzlawick though so I definitely prefer him (although he clearly is based on the first).

  2. @Syl
    Oh thanks, book suggestions are always welcome :) Will check those out when my current pile of book needs restocking!

  3. Thank you for articulating so beautifully my very thoughts regarding Kuhn's book! I have to read it for a graduate course, and while I love science and its history, this book is dreadful. Dry, erudite language coupled with miniscule font makes for a most unpleasant reading experience. But, as you point out, there are small rewards here and there, so I keep reading for those delightful bits of insight that are written with some humanness. I appreciate your review, as I was beginning to believe that I alone felt this way about so popular a book.