An ocean wave
Is it bad form to criticize a review of a book you have not read?
A distinguished psychologist published a scathing review of Mark Turner's new book The Origin of Ideas. You can read the review here:
Abstract Concepts in Science and Psychology
In it, the reviewer criticizes the book for not being scientific. For example, Turner is criticized for describing things in the mind such as "webs" and "scaffolds." The reviewer says "...Turner violates just about every rule of good science: abstract concepts are treated as though they are real things;..."
This statement gets at the heart of a debate in the philosophy of science. Are abstract things real?
When most people describe what it means for something to be real, they say that it has to be made of matter (or, if they are more sophisticated, matter or energy). This is a philosophical view that has many articulate and intelligent adherents. But like many matters, there are also smart and articulate people who disagree.
I believe in something called "scientific realism," which is the belief that if science uses a concept productively to make accurate (enough) predictions, then it is rational to believe that that concept is real. This also sounds like common sense, but people can sometimes get uneasy with the consequences.
Here's a little quiz. There are no right or wrong answers--well, let me rephrase that: there might be right and wrong answers but we can't tell for sure what they are! In any case, which of the things in the following list do you think are real?
- a particular brick
- a particular neuron
- a memory
- the play Hamlet
- a point in a baseball game
- a marriage
- a gust of wind
- a wave in the ocean
Your answers to these questions can give a first-pass idea of what kind of philosophy of science you believe in.
Let's say you thought that a wave in the ocean is real. What is it made of? The obvious answer is water, but the thing about waves is this: they are actually abstract objects that are made of different water at every moment. In what sense is a wave a mile out to sea the same wave that crashes on th beach? Certainly not because they are made up of the same atoms: they are not.
Yet scientists can measure waves, and we can predict what they will do (to some level of accuracy.) As a scientific realist, I am happy to say that a wave is real for this reason.
Abstract concepts are no strangers to psychology. Is a marriage real? Being married has predictable, measurable effects on people (on average, of course.) The behaviorist movement tried very hard to eliminate them completely from psychology, using similar reasoning that our reviewer used. But ultimately psychology could not get very far without talking about beliefs, memories, desires, plans, and goals. If we are to limit "real" things to only those things made of matter, then most of the things we talk about all the time are complete fictions. This doesn't sit well with me.
Is the play Hamlet made of ink? If you want to study Hamlet should you start by studying the physical properties of ink? What if it's on an e-reader? Is it a different play because it's made of different materials?
The reviewer might not like the abstract entities that Turner advocates, but it is unfair to criticize Turner for using abstract concepts at all. Even the reviewer uses abstract concepts in his own work, such as happiness, maturity, and affection. These are abstract concepts too. If they are better concepts, they are better because they can be used in predictions more effectively, but they are not better because they are less abstract.
An Unfalsifiable Theory?
The reviewer also criticizes Turner for having an "unfalsifiable" theory. What does this mean?
Philosopher of science Karl Popper came up with this useful idea. Let's say somebody comes up with a theory. It's fair to ask what kinds of observations one could theoretically make that would show that theory to be false. If people fail to make those observations, then it's great for the theory. For example, if your theory says that all swans are white, then you would have a falsifiable prediction: that nobody would ever find a black swan. The white swan theory is falsifiable.
People differ on how much they like Popper's theory, but I find psychologists to be particularly taken with it. However, it's important to realize that in the history of science, there have been lots of examples of unfalsifiable theories that were productive for future good science, and many that eventually became falsifiable. Should we have stomped those ideas out right away because they were unscientific? No. Ideas can be stepping stones to better ideas. Some good ideas look worse than the competition at first. They need time to mature. (Hofstadter's Godel, Escher, Bach and Minsky's The Society of Mind are both influential, purely theoretical books that suggest theories that are not, at first blush anyway, falsifiable).
Currently string theory in physics is unfalsifiable. It has been criticized for not making enough empirical predictions. However, it is a mistake to say that string theory is not physics research. In fact, some grumble that it's taking over university physics departments. The reviewer says that "the book contains no content that a biologist or physicist would consider `research' at all."
I'm not saying that Turner's theory is right. I haven't read the book. But some ideas are valuable not because they are rigorously tested, but because they will be inspiring for future ideas. Ideally all scientific ideas will be tested eventually, but to limit our thinking to ideas we can test right now is too stringent a criterion, and too dependent on our current technology and imagination.
Pictured: An ocean wave. (Or maybe nothing real at all. You be the judge). From Wikimedia Commons.
Jim Davies is author of Riveted: The Science of Why Jokes Make Us Laugh, Movies Make Us Cry, and Religion Makes Us Feel One with the Universe, which will be released August 5, 2014 by Palgrave Macmillan, and will probably get some reviews that read like the one discussed here.
Read more of my thoughts on falsifiability here: