hard-soft science division

[Squawk 1200]

Do you consider this division to be valid?

 

Math, physics and chemistry are usually considered to be "hard" sciences while sociology, economics, law and political science are considered to be "soft". Anectodal evidence says that hard sciences are more intellectually demanding, harder to master and at college level they have entry requirements and higher dropout rates - does data back this claim up?

 

I have even seen economics being classified as a pseudoscience... by people with academic background in economics (one of them also had a minor in pure math).

 

 

[Delta1212]

I think relative "hardness" is a cultural phenomenon more so than anything inherent to the subject matter of any given field. It's mostly about the degree of rigor which has been applied to the most current theories.

 

For some fields, it is simply more difficult to apply that rigor for a vareity of reasons both practical and ethical. Psychological experimentation is obviously very restricted in terms of the types of experiments it can perform and who it can perform them on, and sociology deals with events that it is difficult or impossible to develop truly adequate controls for.

 

A lot of the softness is down to historical accident as much as anything, though. There was a period when the roots of what would become the field of chemistry was extremely soft in Europe. Similarly, astronomy has a similar problem to a lot of the softer sciences in that it is difficult to set up controlled experiments to explore a lot of phenomena, but it has a rich history of developing mathematical models based on direct observation.

 

Economics, by contrast, has traditionally been more of a mathematical philosophy than really evidence based.

 

I think a lot of soft sciences have been getting harder, but especially when direct, controlled experimentation is difficult or impossible, it takes a lot of time to build up the robust knowledge base that the harder sciences enjoy.

[Squawk 1200]

What's the practical usefulness of less hard sciences?

[Delta1212]

That seems a bit like asking what the practical usefulness of a car with a flat tire is. I mean, technically you can still travel with it, though it's probably a bad idea to try to go any kind of distance, but the real issue isn't that the car overall isn't useful. It's that the tire needs to be patched or replaced before you can use it properly.

 

There is no advantage to a science being soft over being hard, but there is also nothing that forces a science to be inherently soft. There are some where real progress will take a lot of time and effort and in the meantime we'll have to acknowledge a lot of major gaps in knowledge even for ideas that have been preliminarily tested. This is unsatisfying and making potentially unwarranted extrapolations in order to expand the range of what we know into areas that are simply very difficult to test directly is a big reason why certain subjects have remained very soft.

 

We can't tinker with the economics of countries or the world to see what happens nor with the cultural practices of various societies or their political structures and ideologies. We have to go where the light is currently shining and try to develop models based on that. It's neither easy nor especially accurate with the amount of data we currently possess and are capable of collecting for such complex subjects, and as a result, some people in these fields have historically retreated into patching the holes with speculation or, as stated, unwarranted extrapolation.

 

The more we strip away the unfounded assumptions that a lot of these subjects were initially built upon and get more and more data with which to build real models of behavior, and most soft sciences really do have to do with human behavior simply because it is an extremely complex subject where direct experimentation is either extremely difficult or considered unethical, the harder these subjects will become.

 

It also doesn't help that most of them are relatively young compared to the broader categories of most harder sciences and have had less time to build up the bodies of knowledge that it takes to make both robustly accurate and satisfying predictions about the way things work. That last shouldn't really matter, but it does and is a big part about what leads people to make leaps instead of following the data.

 

As humans, we always want an answer, and if the right one is completely unattainable, we'll often settle for making one up.

[Squawk 1200]

My are aof study is law. Should I then be worried about the value of my knowledge in everything except making money?