On the Art of Writing – and Science

I am currently reading Joseph Schumpeter‘s wonderful The Theory of Economic Development. It is wonderful not only in substance, but in style. Science simply isn’t presented in such beautiful prose anymore.

The tender softness of Schumpeter’s writing brings to mind a sentence from Murray N. Rothbard‘s critical comments on Samuel Edward Konkin III’s New Libertarian Manifesto. Rothbard seems to disagree with Konkin on most of his points, but he still welcomes Konkin to partake in the discussion:

“…especially because he cares deeply about liberty and can read and write, qualities which seem to be going out of style in the libertarian movement.”

The last point, that Konkin “can read and write,” is interesting. Rothbard obviously had the same feeling in the 1970s as I get when reading Schumpeter. The context is different, but both of us identify the decreasing quality of writing. How come the skill to write, i.e., to use the language to communicate knowledge and feeling in an easy-to-understand and attractive text, has not only gone out of style – but is also beyond the reach of contemporary writers and scientists? (I will here discuss mainly scientific writing.)

Part of the reason good writing is no longer a desired quality, I claim, is due to over-dividing of sciences into fields and subfields. Part of it is also the faulty belief that real science must necessarily be mathematical.

Division of science

When Adam Smith wrote about the division of labor and how the individual actions of people doing essentially different things come together to create change in a unified direction as by “an invisible hand”, it made sense. In rural society, where most people were working on the family farm, dividing chores meant increasing efficiency. This is even more true for the industrialization, where the division of labor could increase productivity by enormous numbers.

But does this mean division of labor always makes for increased production? Certainly not. History proves only the fact that division of labor was more productive (or, maybe, less unproductive) than the immature market production processes available at that time.

The same lesson can be learned from science, which used to be a field for thinkers who proudly tried to find an all-encompassing theory to explain everything, natural and social sciences alike. These theories were necessarily abstract, but the writings of these great men (I’m afraid there were almost no women) throughout history did provide bases for continued research and thought.

Around the time of the Enlightenment science started to become specialized. Scientists, who were previously nothing but philosophers (in the broad sense), specialized in different fields, and the differences between these fields started to become apparent. Physics, biology, and chemistry became different fields of study, as did economics, political science, and history.

The division of sciences, just like the division of labor, acted as a catalyst, providing focus for the individual scientist (or group of scientists) and thereby elucidating the means for identifying and studying specific phenomena. In this sense, productivity in science increased – but so did the differences between the fields.

One might think that differences between fields of study aren’t necessarily a bad thing. This is true, of course. But it is also true that the fields of study (or disciplines) are arbitrary divisions of science made by scientists to create an environment where one’s work is not “disturbed” by competing theories and worldviews. Also, the differences between disciplines were set a couple of centuries ago – it isn’t obvious that the fields of study would be the same if science was divided today. So a problem might be that different disciplines are implicitly competing with each other.

Also, the different disciplines now have different history and different “gods” (thinkers and scientists considered important in the individual field of study). This means there is a great difference as how economists think of a certain social phenomenon and the way sociologists or political scientists think of the same phenomon. Is this a good thing?

It could be a good thing, but only if the disciplines are open to other ways of thinking, i.e. that they embrace the opportunity of using other perspectives and theories in their work. But this is hardly always the case. Rather, we are seeing a lot of territorial battles between the disciplines even though there are some inter-disciplinary research projects.

Being specific could thus be a problem – sociologists lack the economists’ view of the world and are therefore unable to make use of this perspective. And vice versa. The real problem here is that most fields of study have already done quite a lot of specific research – the scientists know their fields, even though there are quite a few problems yet to be solved. What is needed is consolidation, science as a whole cannot go forward without getting rid of the artificial borders between disciplines and sub disciplines.

And since science tends to be less general and more specific, this also affects scientific writing. Each discipline has its jargon and way of expressing things, which effectively keeps non-scientists as well as scientists of other disciplines out of the discussions. Also, the problems get smaller and more detailed, and thus the study of them takes less notice of the context. Scientific writing thus becomes very technical – and hard to read.

Math as the [only] language of science

The last point made in the previous section is important: science has become very technical. Studying the details rather than getting the big picture requires a use of language that is very specific and unambiguous. This is, of course, a good thing – there is no reason science should be vague and ambiguous. However, this eventually (and, indeed, it has) leads to scientists disqualifying language as a means for communication – instead, math is used as the only possible way of describing and explaining a phenomenon.

Math has a lot of advantages, no doubt: it is clear, unambiguous, and it is expressed in essentially the same way no matter what is your native language. But being clear and “pin-pointing” is also problematic – especially in the social sciences. When one studies the social phenomena, i.e., humans acting, one cannot be mathematically exact. There is no way one can describe the actions of a human being in such exact terms – the value of exactness thus becomes a vice, since it leaves out part of the picture and makes the scientist unable to i) describe the width and depth of the situation, and ii) communicate understanding (cf. Weber’s terms Verstehen vs. Erklären) of the phenomenon.

Also, and this is a sad truth, a scientist used to the mathematically exact language will find him- or herself unable to study the real situation. If the language used for absorbing knowledge is mainly mathematics, then it is no longer possible to interpret the why’s, how’s and what’s. The value of social science, an extremely important part of scientific research, could thus be decreasing simply because social scientists wish to be deemed more “scientific” through using mathematics.

On the writing of scientists

What has been mentioned in rather vague terms in the previous two sections have implications for how science is presented. The “old style” essayist writings of great scientists does not only tend to go away – it might become extinct. Many scientific journal articles today are too concise and too technical to make sense – many theories presented contradict common sense, and sometimes have nothing to do with the real world. The latter is true even though the theory presented is based on real world data!

The scientists are simply blinded by their desperate need to make a clear case, that they forget (or are unable) to ponder on whether it makes sense. The big picture is necessary for detailed science – but the opposite isn’t necessarily true.

So what we have here is a scientific community that has been overly divised and that is self-enforcing the differences between the disciplines at the expense of science. This does not, of course, make sense – but in order to identify this fact one needs to shift perspectives for a minute, from the microscope to the binoculars, and ask the question in real language: does this make sense.

The “old style” scientific reports were not as detailed and did not use math as extensively. That is not necessarily a bad thing – one can make very rigorous scientific research, even math-based, without having to present every detail in a scientific paper. Good research generates new theories to explain and understand phenomena – the theory itself doesn’t have to rely on mathematically exact language. Actually, any theory about a social phenomenon is probably better if it has been developed through good research (perhaps using math), but is stated in general terms – and thus could, and should, be discussed using language and sensible examples.

General theories provide a great basis for further research: such theories are easier to dispute and prove wrong, and they are easier to apply on different sets of data – and they are communicable to out-of-the-discipline scientists. What we need is a revolution in science, a backwards revolution. We need to start writing again, and combine rather than divise.

This is what is so great with Schumpeter in the work cited above: he combines knowledge of different scientists, presumably from different disciplines, and he presents it in common language – he provides a combined scientific theory that people can understand. And thus: he provides the means for further research – and automatically welcomes any attempt to prove him wrong. In other words, he provides good science.