2020-05-04: Archimedes and the sphere

2019-05-16: Glow worms return

2019-04-11: Original memetic sin

2019-01-31: The theory of weight

2018-11-06: Origins of telephone network theory

2018-10-24: Modern thought

2018-09-10: Feeding a controversy

2018-06-11: Glow worm distribution

2018-04-23: Outlawing risk

2017-08-22: A rebuttal on the beauty in applying math

2017-04-22: Free googles book library

2016-11-02: In search of Theodore von Karman

2016-09-25: Amath Timeline

2016-02-24: Math errors and risk reporting

2016-02-20: Apple VS FBI

2016-02-19: More Zika may be better than less

2016-02-17: Dependent Non-Commuting Random Variable Systems

2016-01-14: Life at the multifurcation

2015-09-28: AI ain't that smart

2015-06-24: MathEpi citation tree

2015-03-31: Too much STEM is bad

2015-03-24: Dawn of the CRISPR age

2015-02-12: A Comment on How Biased Dispersal can Preclude Competitive Exclusion

2015-02-09: Hamilton's selfish-herd paradox

2015-02-08: Risks and values of microparasite research

2014-11-10: Vaccine mandates and bioethics

2014-10-18: Ebola, travel, president

2014-10-17: Ebola comments

2014-10-12: Ebola numbers

2014-09-23: More stochastic than?

2014-08-17: Feynman's missing method for third-orders?

2014-07-31: CIA spies even on congress

2014-07-16: Rehm on vaccines

2014-06-21: Kurtosis, 4th order diffusion, and wave speed

2014-06-20: Random dispersal speeds invasions

2014-05-06: Preservation of information asymetry in Academia

2014-04-16: Dual numbers are really just calculus infinitessimals

2014-04-14: More on fairer markets

2014-03-18: It's a mad mad mad mad prisoner's dilemma

2014-03-05: Integration techniques: Fourier--Laplace Commutation

2014-02-25: Fiber-bundles for root-polishing in two dimensions

2014-02-17: Is life a simulation or a dream?

2014-01-30: PSU should be infosocialist

2014-01-12: The dark house of math

2014-01-11: Inconsistencies hinder pylab adoption

2013-12-24: Cuvier and the birth of extinction

2013-12-17: Risk Resonance

2013-12-15: The cult of the Levy flight

2013-12-09: 2013 Flu Shots at PSU

2013-12-02: Amazon sucker-punches 60 minutes

2013-11-26: Zombies are REAL, Dr. Tyson!

2013-11-22: Crying wolf over synthetic biology?

2013-11-21: Tilting Drake's Equation

2013-11-18: Why $1^\infty != 1$

2013-11-15: Adobe leaks of PSU data + NSA success accounting

2013-11-14: 60 Minutes misreport on Benghazi

2013-11-11: Making fairer trading markets

2013-11-10: L'Hopital's Rule for Multidimensional Systems

2013-11-09: Using infinitessimals in vector calculus

2013-11-08: Functional Calculus

2013-11-03: Elementary mathematical theory of the health poverty trap

2013-11-02: Proof of the circle area formula using elementary methods

The theory of weight

In looking back over the history of applied mathematics in my recent studies, I was reminded that the theory of statics and torques actually predates the Renaissance period considerably. Archimedes, in two of his surviving works, exhibited some sophisticated results on centers of gravity and equilibria of floating bodies circa 250 BC! His axiomatization of the "law of the lever" is the simplest of these, coming just a generation after Euclid's monumental work.

Even before one can ask about the relation between "weight" and length, it is an interesting question to ask how one extends the concepts of Euclidean geometry to account for a quantity that can not be described soley in geometric terms like volume. Clearly, the practical concept of weight is very old, and can be found in the form of scales in surviving Egyptian works circa 2000 BC. And it may have been much earlier, since the Egyptian standardization of the cubit may have occurred around 3000 BC. But why is it reasonable to believe weights of different substances will behave comparably? Why is it OK to weigh food using stone weights or lead weights? Clearly stone and lead are different substances, and early science provided very poor theories of what this difference actually was. Might not the differences mean that scales using stone weights and scales using lead weights were fundamentally incompatible?

When you consider this question, the mythological use of scales at the gateway to the Egyptian afterlife to distinguish the "light hearted" from the "heavy hearted" takes on a surprising substance. An ancient illustration of this is often referenced in a certain class of sweeping historical documentaries as a reference for the ancient perception of humanness (For example, "The story of science" ep 6). A person's heart, representative of the accumulation of good and bad deeds, would be weight on a scale against a feather, with the balance determining their fate. The scale would have been a sophisticated piece of technology 4 thousand years ago, yet one that had already entered consciousness to a point where it played a central role in the most important moment in mythology. Yet the way it was employed, balancing things that seem so different, suggests that its operation retained a mystical quality. It was important that all witnesses could see and recognize what things were being compared, and thence that the outcome was the embodiment of some supernatural will of the universal, and even the gods were subject to its will. Not at all what we think of today when we use and trust scales. (this same issue would raise it's head again in 1299 Florence, for example, when the use of Hindu numbers was banned in some places in favor of the familiar and transparent abacus with Roman numerals.

Which brings me now to the main point of this post -- the axiomization of the concept of weight, in alignment with that of classical geometry. It remains a bit of a mystery to me to what extent the concept of weight is an empirical concept and to what extent it is a Platonic abstraction like the point and line. The modern convention is that we define mass, rather than weight, and that (atleast was until recently) mass is compared against a cubic centimeter of water at 4 degrees Celsius. But rarely does it seem to be mentioned why this convention is actually justifiable -- why is it we can adopt one universal standard for the measurement of weight (and mass). Might there not be properties in weight that can not be measured with one such comparison?


Suppose we have weights of two different substances (say, wood and stone), and a scale with two arms that can be used to test balance. These arms are not necessarily of equal length, but their lengths will be constant and we will always weigh wood on the left and stone on the right.

We can think of the scale as a mechanical implementation of a function that takes two inputs, some volume of wood on the left and some volume of stone on the right, and outputs which of the two weights more -- the wood or the stone. We can then draw a picture, of the space of inputs coloring balance tips towards wood green, and balance tips towards stone purple.

Produced by GNUPLOT 4.4 patchlevel 3 Volume of stone Volume of wood Tips toward stone Tips toward wood

The boundary between the green and purple regions will be special because for any volume inputs on the boundary, the scale will be balanced -- the volume of wood will weight the same as the volume of stone. What shape will this boundary take?

We can imagine this boundary having some shapes but not others. For example, if the scale is tipped toward wood, and we add more wood, it should remain tipped toward wood. And if the scale is tipped toward stone, and we add more stone, it should stay tipped that way as well. These are axioms based on our real-world experiences with weight. Archimedes (and probably many before him) observed these axioms as well and used them. Also, if there is no wood, the scale can not tip that way, and if there is no stone, that scale can not tip that way either. If both sides of the scale are empty, it is balanced. It follows then that the boundary between stone and wood (green and purple) must start at the origin, and divide the input space into exactly two pieces, one green and one purple (there cannot be any "bubbles" of color in the picture). The exact shape of the boundary, however, remains unknown. It could be a parabola, or a line, something wavey -- all of these are consistent with our elementary axioms.

To go further, we will produce a theorem, and consider its implications. Let us define the function \(F(x)\) as the volume of stone needed to balance a volume \(x\) of wood. And suppose there are volumes of wood \(a\) and \(b\) such that the sum of the volumes of stone needed to balance each of these is less than the volume of stone need to balance their sum, so \(F(a) + F(b) < F(a+b)\). Then we can show the following.

Theorem: Assuming we have two equal scales next to each other, and we can move weights between these scales without any work when they are tipped the same way, then we can lift as much stone up to the height of the scale as we want without doing any work.

Proof: First, by the continuous divisibility of volumes, there are volumes of stone \(c\), \(d\), and \(e\) such that \(F(a) < c\), \(F(b) < d\), and \(c + d + e < F(a+b)\). Now, suppose we set up two identical scales side-by-side. To start with, we weigh \(a + b\) against \(c + d\), and the scale is tipped to wood. Now, split the weights so we weigh \(a\) against \(c\) and \(b\) against \(d\). Both scales now tip in toward stone. Now, me move all back to a single scale and add the extra stone volume \(e\). The scale now tips towards wood, and we can offload the extra stone volume \(e\) as work done, return us to our original scale state, but with more work done lifting stone. We can repeat this cycle as many times as necessary to lift any designed about of stone. QED

This theorem is, in spirit, a perpetual motion theorem. Once the stone is lifted, it can be used to do other work, and hence get things done. We then may ask ourselves, "Do we believe in perpetual motion machines?"

If you do not believe in perpetual motion, then it must be that for every pair of wood volumes \(a\) and \(b\), \(F(a) + F(b) \leq F(a+b)\). If we believe the same argument holds for stone lifting wood, then \(F(a) + F(b) \geq F(a+b)\), and hence, \(F(a) + F(b) = F(a+b)\) for all \(a\) and \(b\).

This functional equation's solution is a line, so \(F(a) = k a\) for some constant \(k\). In modern language, we'd call \(k\) the relative density of stone compared to wood.

We can repeat this argument for every other pair of substances, like lead, gold, and water, and define relative densities for each of these comparisons.

It then follows that all weights can be calculated using a standard comparison like water, and from here on, we need only consider weighting one standard against another. We can now move on to determining the law of the lever.

This seems like a useful axiomatization of the concept of weight to me.