We should also have a SI unit for degrees of certitude, which would be useful for quantifying the discussions over in Great Debates. I propose that it the basic unit be called a certimeter. It would probably be a logarithmic scale, that is, each additional degree of certimetry indicates ten times the certitude of the previous degree. As such, the certimeter measure of an argument or belief could be negative or positive, on a scale of (theoretically) -∞ to +∞.
This is way outside my scope of understanding, but I am wondering if it’s because absolute zero is not really “zero energy.” Based on what I’ve read, if an atom or molecule is at absolute zero, it still has a finite amount of kinetic energy.
OTOH, mass is based on just one point (as of 2014): the International Prototype Kilogram (IPK). I supposed you could say it’s based on two points: no mass has a defined mass of 0 kg, and the IPK has a defined mass of 1 kg. But that would be kinda silly. So perhaps defining absolute zero to be 0 K is equally silly, hence your point.
Can you elaborate/explain this?
Yes, I truly made an extremely excellent point there.
But then this kind of shot the whole thing through the heart (from the wikipedia article on “thermodynamic temperature”):
"By international agreement, the unit kelvin and its scale are defined by two points: absolute zero, and the triple point of Vienna Standard Mean Ocean Water (water with a specified blend of hydrogen and oxygen isotopes). Absolute zero, the lowest possible temperature, is defined as being precisely 0 K and −273.15 °C. The triple point of water is defined as being precisely 273.16 K and 0.01 °C. This definition does three things:
It fixes the magnitude of the kelvin unit as being precisely 1 part in 273.16 parts the difference between absolute zero and the triple point of water;
It establishes that one kelvin has precisely the same magnitude as a one-degree increment on the Celsius scale; and
It establishes the difference between the two scales’ null points as being precisely 273.15 kelvins (0 K = −273.15 °C and 273.16 K = 0.01 °C)."
Allright, I read a bio of Kelvin and a couple books on thermal physics and writings from Clausius and Carnot and god knows who all else plus a bunch of other references that I thought made consistent sense, to the effect that you don’t have to have a zero to define a unit and don’t need a scale for absolute temperature, and then there’s this in Wikipedia (a source I usually find really good for technical stuff like this). So, this isn’t a very satisfying situation. I want to dig around some more. It’d be really nice if I can find something that says there’s no need to define the zero end, and says so in such a way that everybody agrees it simply must be so. Uh… stay tuned?
But the point should still remain that you can create a relative scale that guides the assignment of numbers to some physical phenomenon, and scales so defined have no requirement that they be in a real unit and be proportional to the true measure of the phenomenon. And, that having a real unit is more meaningful. And, finally, that either no scale needs a zero to define it, or EVERY scale needs a zero (or at least some second point) to define it.
The point there, which is in no way invalidated by the spot of ugliness about zero definitions, is that any tool that gives you a unique reading for every temperature you expose it to can be the perfectly valid basis for a temperature scale. For example, elemental metals (though not necessarily alloys) generally have electrical resistances that are higher at higher temperatures and lower at lower temperatures. You could create an ohmmeter with a metal probe, and then mark its scale with the numbers from zero to a thousand, and declare that the Bloom temperature scale, and you could use it for most of the things we measure temperature for. Or for example you could make narrow glass tube with a bulb on the end and put liquid in it (this actually DID lead to temperature scales IIRC). However, you are taking advantage of some thermally dependent behavior that may or may not be a linear function of thermodynamic temperature (and in fact my examples are not linear functions, they’re just close).
Likewise, you could point a bit of tubing out the front of your car and connect it to a pressure gage and divide the gage scale into 300 increments and make your own speedometer (for this example I’m assuming there’s no wind and the air temp and barometric pressure aren’t changing). However, this won’t be linear with speed (in fact it will vary as the square of speed). So, you will find that when you “double” your speed according to this speedometer, it won’t take you half the time to get someplace, it will take about 71% of the time.
Long story short, there’s an underlying meaning to thermodynamic temperature that goes deeper than anything thermometers do. Well, most thermometers. There are thermometers based on Johnson noise in electronic devices, and also thermometers based on gas density that use extrapolation techniques, and these actually give thermodynamic temperature, but they’re so hard to use nobody uses them except for the people who explain what other kinds of thermometers are actually doing.
Thanks much.
Now explain why only some amps can go to eleven.
+1 
Agreed.
Another thing that bothers me (about using absolute zero as a reference point) is that it can’t be achieved; unlike the TP of water, I can’t go out and buy an absolute zero reference cell. Even if one were available, would you really want to stick your $7000 SPRT into it? (It would be way outside its temperature range.) One wonders why they didn’t use two practical fixed points (e.g. the TP of water and the FP of aluminum) for defining the practical temperature scale. I would assume it has something to do with meshing the thermodynamic temperature scale with our current practical temperature scale. But again, that’s way outside my area of understanding…
Well, I think there are other realizable statements that would go along with the idea of absolute zero. Radiated energy for a blackbody will only vary as T^4 with the right conception of where absolute zero is. The ideal gas law would only work right with the same right conception. Ideas about entropy and the number of possible states of a system, likewise - so specific heats would only go to zero at the proper limit. And so on. If it was defined in terms of two realizable temperatures per se, then the various limits of testable things as T goes to a true zero would not work right. The fact that there’s a concept of a true zero here could be the basis of a definition, I’d think, even if it’s defining a point.
What would we call one tenth of a certimetre?
Nevertheless, the zero point is the one actual, non-arbitrary reference point for Kelvin. The 273.16 figure is a retroactive “definition” based on counting Celsius-magnitude degrees from zero to the triple point. Right?
Right. And we could have defined the triple point as exactly 100 kelvin, or 1000 kelvin. At which point we’d have to give up the convenient idea of Celcius degrees and kelvins being the same size. That convenience outweighed the potential advantages and logical cleanliness of making kelvins some other size.
Yup.