Hacker News Reader

Comment by rini17 | Hacker News Reader

The elephant in the room is that humanities have zero predictive power. And reaction to unexpected results is markedly different, too. If the eclipse did not happen exactly as predicted, hard sciences would (mostly) excitedly scramble to understand how and why. If Trump gets elected...you get stuff like this article "oh my what kind of person we are in this Trump era".

View on HN

I suspect you strongly misunderstand what humanities provide: a rounding out of one's expertise and an appreciation of the qualitative world.

You seem to presume predictive power is the preeminent criterion — the only thing worth optimizing for.

Author seems to disagree with you - he basically says humanities should provide better input to political discourse.

Also, it's the author that started using comparisons with hard sciences. In this context I have used the "elephant in the room", something to be aware of, not necessarily to optimize for.

Author may also be wrong.

The humanities, or rather, the Liberal Arts from which they originate, are not about predictive power as such.

Rather, they're about information processing, and sets of useful priors.

The fundaments of the seven Liberal Arts are the trivium and the quadrivium. (Trivium => trivial, by the way.)

The three elemnts of the trivium are grammar, logic, and rhetoric. Or, to put that in an information-technology / information-theoretic frame: input, processing, and output.

The four elements of the quadrivium are arithmetic, trigonometry, music, and astronomy (or astrology). Or: quantity, quantity in space, quantity in time, and quantity in space and time. Possibly an early development of systems dynamics.

I'm not arguing that present humanities have strayed far from these roots, or that they're not exceptionally incenstuous and inward-focused. They have and they are.

But your critique that they are not of themselves predictive has been launched at the wrong target. You misunderstand the domain. (Don't feel bad, many do, and I've only recently realised the error myself.)

More: https://redd.it/6x7u6a

Descriptive science is equally important as predictive science, but either way, science is not our only tool for knowledge. Statistical predictive models do happen to be in vogue, though.

> Descriptive science

The whole point of science is for it to be predictive though. Calling "descriptive science" (whatever that is) a science is sort of... wrong!

The point of mathematical models is to be predictive. But I wouldn't call prediction "the whole point of science". That's reducing science to a means-to-an-end.

I think science is more commonly understood to be about discovery and exploration? Although it is commonly applied and turned into engineering, expansion of knowledge is still an end in itself.

Ever heard about Darwin's "Voyage of the Beagle"? The purpose was to collect specimens and chart coasts -- purely descriptive. Even the idea of predictive natural science was a little awkward in the face of the religious concept that God created the world as-is/will-be.

The elephant in the room is that humanities have zero predictive power

Are they meant to? I don't read Tolstoy or play the piano because I think it will enable me to predict the future.

Arts are not meant to have predictive power. Sciences should have predictive power. The study of literature and music is the study of art. If psychology or sociology currently have little predictive power, they should be called arts instead of sciences. Falsifiability is the corner stone of science.

> The elephant in the room is that humanities have zero predictive power.

An absolute statement. Do you really think it's true?

Are you really saying you can gain no predictive power by studying history? Ethics? Sociology?

> If Trump gets elected...you get stuff like this article "oh my what kind of person we are in this Trump era".

Science did a great job predicting Trump would be elected.

There are generally accepted scientific theories that have zero predictive power. The theory of evolution is a great explanation of how we got here. But it can't predict what's next.

That's not true at all. Yes, you can say that the theory of evolution can't predict what man will be like in the year One Million or other pop culture ideas of how evolution works, but you can make the same claim about physics -- you can't predict what a given particle will be doing then either.

But there is lots of predictive power in evolutionary theory. The core of modern evolutionary theory is population genetics which is as mathematical as physics and can make predictions of whether certain alleles will increase or decrease in a population, for example.

Of course it can. We can apply selective pressure to a population and hypothesise the likely outcome. You can do this on a small scale in a Petri dish by selecting for antibiotic resistance, for example. And then you can validate the hypothesis by sequencing to determine the exact genetic change which led to the outcome.

If you wanted to predict the future evolution of a more complex species, that would be harder. Because we don't yet know what the selective pressures in the future will be, and the timescales involved are much longer. But if you define a set of selection criteria you can absolutely test a hypothesis by applying them today.

All scientific theories should have predictive powers, otherwise they wouldn't be falsifiable, and, as such, not science.

Sometimes, as with your example, it's less obvious, but say, if a new species is discovered, evolution predicts there will be genetic differences to its closest cousin or ancestor and that the differences will be related to its fitness to thrive within a particular environment.

Also note that "predictive" doesn't need to be about future events. It can also mean new discoveries within existing records (like a feathered dinosaur).

Evolution is probably the wrong example. I think a better example would be something like inflationary cosmology.

At least one reason to study humanities is to study philosophy of science, and understand how new and insufficient an idea like "science must have predictive power" is.

Reply on news.ycombinator.com