Comment by ycombinete | Hacker News Reader

Comment by ycombinete | Hacker News Reader

I haven’t read this paper, but instinctively flinch at any kind of dismissal of the past.

I would think this comes down to how we define philosophy. On scientific matters sure Aristotle likely has little to offer; but human nature has not changed at all since Plato, or Siddhartha we’re born.

After saying the following I feel that it would be hard for them to argue their point very strongly. So I’m definitely going to read the paper:

To be perfectly clear: my claim is not that we should not be doing history of philosophy. There are all kinds of reasons why reading and talking about the Critique of Pure Reason or the Republic are worthwhile: studying these seminal texts is an inherently interesting intellectual pursuit; reading them is often tremendously enjoyable; and familiarity with these texts can be very valuable to intellectual historians for the insights into culture, knowledge and morality they may contain. There are thus many excellent reasons to engage with the history of philosophy. Gaining traction on the aforementioned philosophical problems, however, is not one of them. This means that I am not arguing against historians of philosophy and what they do, but against what could be called philosophical historicists, that is, those who seem to think that at least one good method of thinking about knowledge or justice is to study what historical authors have written about knowledge and justice a long time ago. This, I argue, is a mistake.

Do you believe someone needs to understand the history of math to effectively use math?

I don't think anyone is claiming historical context isn't helpful, but these are tools for thinking. This doesn't go anywhere near the 'history repeating itself' mantra. The only way it could get close to that is for someone who is attempting to discover new math (or philosophy), but that's a very small subset of those who benefit from these tools.

Yes. So many things in math, in science, in programming, only make sense if you understand where they came from and in what context. Otherwise things just seem weird and arbitrary and hard to understand.

I don't need to understand how a fence post evolved in order to successfully use fence posts.

To use them well I think you would. People who haven't been exposed to a history of a method or technology often say "I think this is silly; I can think of a better way to do this". Sometimes, extremely rarely, they are right. But more often, what they are thinking of is something that was tried and was found not to work well. Knowing the history of what you are doing, and what things worked and failed prevents you from making the same mistakes your forerunners did.

There's always an odd dichotomy between those who do, and those who do not.

No farmer in existence would ever agree they need to understand the historical context to successfully use a fence post, or that they need it to be able to critically evaluate their specific needs and adjust their usage of fence posts to meet those needs.

I find this attitude amongst software architects a lot as well. The stereotype of the hapless academic didn't come about because there's no truth to it.

In math everything can be derived directly from the underlying axioms given enough time. You don't need historical context. You may want the set of previously proven things to save some time, but you don't need it.

Use? No, math is a tool.

Extend? Yes. Math is a tool that came from a context, so there might be implicit assumptions made on choosing that particular notation and definition that might or might not be justified. Appreciating where and idea comes from lets you work with the idea. I also think it helps learning and generalising.

A competent programmer with no knowledge of history can deal with things like e.g. non-monotonicity in time by just writing robust code. But it helps to know where it comes from: clock drift, user modification, error, leap seconds etc. If you know one of these, you can leverage different assumptions.

You don't need to understand the history of monotonicity to understand and effectively use the concept.

Will it make you better? yes, of course, but that isn't what's being discussed here.

how about you read the paper?

There's a sentence in my comment where I literally say, "I'm definitely going to read the paper."

Reply on news.ycombinator.com