Live data from Hacker News

Harvard 1869 entrance exam

spectrum.columbiaspectator.com

111–120 of 141 posts

Re: Harvard 1869 entrance exam

#111
post #92
post #71

Earlier quoted context omitted.

That works for positive integer bases and exponents, but try giving an "intuitive proof" with irrational exponents. Most things in math, even seemingly obvious things in arithmetic, require a lot of shared background knowledge (at least propositional logic, basic set theory, and a construction of the natural numbers) for two people to even converse formally.

It still works with irrational exponents (start with fractional numbers and work towards that). It also works with imaginary exponents. It works because the power notation is short hand. But that wasn't the point of my reply. The point was that stating things multiple ways assists us in understanding. Does this not match your experience?

It still works with irrational exponents (start with fractional numbers and work towards that).

So why e^pi * e^e = e^(pi + e)? Yes, it follows from the fact that it works for rational numbers, but in order infer this, you'd need to prove the continuity of exponential function, which is nontrivial at best.

Of course, if you define a^x to be the unique continuous function f: R -> R, such that f(1) = a and f(a)f(b) = f(a+b), as soon as you proved the existence and uniqueness of this function, this follows straight from definition.

There are also different definitions of exponential functions, like exp(x) = lim n->inf (1+x/n)^n, or exp(x) = sum_{n=0}^inf x^n/n! . How easy it is to prove now that exp(pi)exp(e) = exp(pi+e) ?

Re: Harvard 1869 entrance exam

#112
post #75

The thing I noticed (sorry, LaTeX fan here) was how well they could typeset math and Latin all the way back in 1869! I have to say I'm impressed, and it makes me a bit sad that most of my college exams looked worse (typographically) than a paper produced over a century ago. Look at those goddamn gorgeously even margins, the ligatures, the kerning of the italics, and the protrusion of the hyphens. Hell, this makes the…

An 1869 MIT entrance exam was posted here a while back and the typesetting was similar:

http://news.ycombinator.com/item?id=1967040

http://bm98.posterous.com/did-they-have-latex-in-1869

Re: Harvard 1869 entrance exam

#113
post #81

Earlier quoted context omitted.

In those days, being an educated person meant memorizing a lot of things, because you couldn't haul your library with you everywhere you went, and in any case it didn't have full text search. Separate a modern educated person from google/wikipedia/etc. and see whether they can still converse at the same lavel. Most can't.

There was also a heavy emphasis on a particular subset of history that made up "classical" studies. Besides the river questions, the history and geography questions were about ancient Greece and Rome. There was nothing, for example, on medieval or renaissance history. I suspect that a comparable modern student would have a broader understanding of ancient history than what was shown on this test.

I suspect not. A comparable modern student would have been exposed to a broader range of ancient history which would also have been shallower and, having been encouraged to "understand" instead of memorizing any of it, would later be left with no real knowledge of anything specific, therefore no real understanding.

Re: Harvard 1869 entrance exam

#114
post #23

The math here is surprisingly weak. I'm not sure about where you guys went to school, but this is fifth grade stuff. Sure, there's a lot of memorization, more classics. But wowzers. No math.

You're exaggerating more than a little bit. There's 2 trigonometric proofs, and a number of less than obvious geometric proofs, especially the latter ones pertaining to the circle. The rational equation in #8 on algebra going to involve solving a cubic. And although #7 in arithmetic wouldn't be too hard if you worked entirely in pence, it is still a trickier problem in the days before decimalization. Granted, it's we…

What age are Americans typically when they take SATS? The geometry stuff is not so bad if you can give visual proofs, pretty hellish otherwise. I think I had to know quite a lot of this stuff for A-level maths as well as Calculus, basic Statistics, Linear Algebra and Applied Maths. But we take these exams when we are 17 and it was 1 of 3 subjects we specialised in for 2 years. Also, regarding your point about calculators, our exam questions were set so that it generally wasn't practical to use a calculator to solve them. E.g. some questions specifically stated that you should give the answer as surd or fraction, it was generally easier / quicker to do the working out manually.

Re: Harvard 1869 entrance exam

#115
post #75

The thing I noticed (sorry, LaTeX fan here) was how well they could typeset math and Latin all the way back in 1869! I have to say I'm impressed, and it makes me a bit sad that most of my college exams looked worse (typographically) than a paper produced over a century ago. Look at those goddamn gorgeously even margins, the ligatures, the kerning of the italics, and the protrusion of the hyphens. Hell, this makes the…

Out of curiosity, how did they typeset those mathematical formulae? Still by carefully tiling little metal blocks or with something more advanced.

Re: Harvard 1869 entrance exam

#116
post #34

> What is the reason that when different powers of the same quantity are multiplied together their exponents are added? As a math professor, I think this is a great question. Students learn that math is about manipulating formulas and equations, or about excessive formalities. But being able to explain simple arithmetic facts in clear and plain English is often neglected, and is of the utmost value.

Who here can come up with the most concise explanation?

Another "proof":

x^n = exp((log x)n). By definition, exp(n+m)=exp(n)exp(m). By definition, a(b+c) = ab + ac (for the log x thing). QED.

By the way, (the infamous calculus book) Baby Rudin has the poor reader show this property holds in exponentiation for reals, starting with integers and via rationals, as an exercise on its first chapter. Insane difficulty for me, even though the author practically holds your hand along the way! Cool read, though.

Re: Harvard 1869 entrance exam

#117
post #58

Earlier quoted context omitted.

That's the point -- you don't understand it well enough to explain it.

I don't think that's the case. As baddox says, it's trivial to show it is true, especially using simplified definitions for exponentiation (i.e. sticking with integer or perhaps rational exponents), but demonstrating truth doesn't tell you about the "reason". Is the question about a philosophical position as to how mathematics relates to God? A "reason" seems to imply a purpose.

I would say that "reason" in mathematics is akin to "motivation" for a definition.

In this particular case, the property a^x a^y = a^(x+y) (plus some very weak technical condition, like Lebesgue measurability) uniquely defines exponential functions.

So, in hindsight, you can think of exponentials as arising in the classification of homomorphisms from the additive group to the multiplicative group of reals.

It actually goes deeper than that. You can extend the reasoning to complex numbers (as everyone knows), to matrices, to Lie algebras, and probably beyond.

Re: Harvard 1869 entrance exam

#118
post #96
post #64

Earlier quoted context omitted.

It isn't "chronological hubris" that the vast bulk of human knowledge has been generated since 1869. Precise measurement is difficult to even define but it's difficult to imagine a non-pathological definition for which that would not be true. Personally I'm a huge advocate for the fact that what one might call "wisdom" is not unique to our age and may indeed be getting a bit lost in the shuffle, but nevertheless, the…

The bulk of current (2011) human knowledge was set in stone a thousand years before 1869.

In what way? In the way that most of physics 'existed' back then, yes. But the vast majority of 'knowledge' (actual explanations on how nature works, mathematics, etc. etc.) was 'discovered' or 'described in detail' over the last couple of decades.

Re: Harvard 1869 entrance exam

#119
post #45

Earlier quoted context omitted.

I don't agree with you. I did 2 years of latin and one year of greek. I don't feel like it really helped me. I actually found it pretty useless except the syntax/grammar part which can be good to understand new languages easier. It's exactly like learning scheme. Seriously who fucking cares about scheme? I see latin and greek in a Harvard test as a part of distinguishing highly educated kids from the others.

> It's exactly like learning scheme. Seriously who fucking cares about scheme? Err... This website is written in Arc (a cousin of Scheme), which itself is implemented in MzScheme. So by extension, you care about scheme, and so does everyone else here.

Right, because all the readers of, say, CNN.com care about whatever language that website is written in?

Re: Harvard 1869 entrance exam

#120
post #80

> Find cube root of 0.0093 to five places of decimals Nice, reminds me of > Find 7th root of 0.9999 to four places of decimals :)

The latter is much much easier: the answer is 1.0000, and "obviously" so. Now, if you happen to know that 21^3=9261 then you can do the 0.0093 one to 5dp with only a few lines of calculation, but it's distinctly more work than the 0.9999 one even so.
Post reply on HN