# Redefinitions of SI units

Page 2 of 3
Goto page Previous  1, 2, 3  Next

1304011.  Tue Nov 20, 2018 12:12 pm

dr.bob wrote:

 GuyBarry wrote: You can only measure relative to a standard. The current standard assumes a constant speed of light. There are very good theoretical reasons for making this assumption but it's still only an assumption.

It really isn't. You mentioned something similar last time we had this discussion, and I'll repeat what I said then:

Physics is completely governed by data. The fact that they've fixed the value of the speed of light is based on decades of experimental data.

I know it is. The evidence that the speed of light in a vacuum is constant is overwhelming.

 Quote: It's not an assumption.

Yes it is. It's an assumption backed up by evidence from decades of experimental data. It's not a mathematical constant like pi whose value can be calculated from first principles.

Why is the speed of light only known to nine significant figures? Because that's the level of accuracy to which we've been able to calculate the speed of light so far. If the speed of light actually fluctuated between 299,792,457.9 m/s and 299,792,458.1 m/s, we wouldn't know about it.

 Quote: And here we have the clear difference between the theoretical perfect world that exists in principle, and the messy, normal world that practical scientists deal with pragmatically every day. As you point out, it was clear that the IPK was losing mass. This was well known despite, according to you, being theoretically impossible.

It was losing mass relative to the other prototype kilograms. It wasn't losing mass under the definitions of the SI. Under those definitions, those other prototype kilograms were gaining mass.

Quote:
And yet they said exactly that. Here's an article from 2011 in which Dr Terry Quinn, Director of the BIPM, is quoted as saying:

 Quote: Measurements over the past century have shown that the international prototype has lost around 50 micrograms

Then he was talking rubbish. The mass of the IPK was 1 kg in 1889, and will continue to be 1 kg right up until 20 May 2019. It is clearly impossible for 1 kg to be 50 micrograms less than 1 kg. No one can change the rules of arithmetic - not even the Director of the BIPM!

1304083.  Wed Nov 21, 2018 6:14 am

 GuyBarry wrote: It's an assumption backed up by evidence from decades of experimental data.

I refer you to the OED:

assumption /əˈsʌm(p)ʃ(ə)n/ [noun]: A thing that is accepted as true or as certain to happen, without proof.
(my emphasis)

I repeat, it is not an assumption.

 GuyBarry wrote: If the speed of light actually fluctuated between 299,792,457.9 m/s and 299,792,458.1 m/s, we wouldn't know about it.

...yet!

Measurement technology improves as time goes on. One day we will be able to measure the speed of light to such accuracy. If we discover that it is fluctuating between 299,792,457.9 m/s and 299,792,458.1 m/s, then we will scrap our theories about a constant speed of light, come up with a new theory that includes a varying speed of light, and redefine our SI units.

 GuyBarry wrote: Then he was talking rubbish.

And here, ladies and gentlemen, is a perfect demonstration of the difference between mathematicians and physicists.

Mathematicians live in a perfect, theoretical world where everything behaves according to the rules. They build theories based on axioms which are assumed to be "self-evidently" true. These are then inviolate and any problems encountered on the way must be issues with other parts of the theory, not the axioms themselves.

In contrast, physicists recognise that the universe is a messy, problematic thing. You can theorise all you like about how it works, but the universe has a habit of turning 'round and behaving in a way you least expect. When that happens, physicists are forced to admit they got it wrong and redefine how they think the universe works.

So, when the international standard of the kilogram appears to be varying in weight, Dr Terry Quinn CBE FRS, Emeritus Director of the BIPM and a recognised world authority in metrology, will happily say "clearly the IPK is changing weight, no matter what the SI definition says in principle." In contrast, GuyBarry, self-confessed mathematician, claims that he's "talking rubbish" because nothing can ever conflict with the perfect world of the SI definition.

 GuyBarry wrote: No one can change the rules of arithmetic - not even the Director of the BIPM!

This is not arithmetic, this is metrology.

This is not maths, this is physics.

1304091.  Wed Nov 21, 2018 7:09 am

 dr.bob wrote: Mathematicians live in a perfect, theoretical world where everything behaves according to the rules. They build theories based on axioms which are assumed to be "self-evidently" true. These are then inviolate and any problems encountered on the way must be issues with other parts of the theory, not the axioms themselves.

That view of what an axiom is is very much a thing of the past. Nowadays an axiom is seen as a defining fact of a (particular) mathematical system. Different systems will have different axioms. For example, in the natural numbers it is an axiom that every natural number is greater than or equal to zero. This is not an axiom in the integers (and indeed, isn't even true).

1304108.  Wed Nov 21, 2018 9:03 am

dr.bob wrote:
 GuyBarry wrote: It's an assumption backed up by evidence from decades of experimental data.

I refer you to the OED:

assumption /əˈsʌm(p)ʃ(ə)n/ [noun]: A thing that is accepted as true or as certain to happen, without proof.
(my emphasis)

I repeat, it is not an assumption.

So where's the proof that the speed of light in a vacuum is constant? There isn't one. As this article puts it:

 Quote: It should be noted that Einstein did not actually PROVE the constancy of the speed of light in all frames of reference. Rather, it is an axiom (an underlying assumption) from which he derived the rest of his theory. The axiom can be experimentally verified, but it is not proven in any theoretic sense.

 dr.bob wrote: If we discover that it is fluctuating between 299,792,457.9 m/s and 299,792,458.1 m/s, then we will scrap our theories about a constant speed of light, come up with a new theory that includes a varying speed of light, and redefine our SI units.

Right. And the fact that you admit of the possibility that the speed of light might not be constant indicates that there cannot be a proof that it's constant. If there were a proof that the speed of light were constant, then it would be impossible to construct an experiment to demonstrate that it is not constant, just as it's impossible to construct an experiment to demonstrate that 0 = 1. You can't demonstrate things that have been proved to be false.

 Quote: Mathematicians live in a perfect, theoretical world where everything behaves according to the rules. They build theories based on axioms which are assumed to be "self-evidently" true. These are then inviolate and any problems encountered on the way must be issues with other parts of the theory, not the axioms themselves.

 Quote: In contrast, physicists recognise that the universe is a messy, problematic thing. You can theorise all you like about how it works, but the universe has a habit of turning 'round and behaving in a way you least expect. When that happens, physicists are forced to admit they got it wrong and redefine how they think the universe works.

And there's nothing wrong with that, of course. But what you can't do is continue with the old definitions and the new ones at the same time. You have to operate within a model that's internally consistent. That's why the SI is constructed in the way it is, with rigorous definitions of base units and rules that relate the derived units to the base units. If you decided to redefine the kilogram, but (say) kept a definition of the newton based on the old definition of the kilogram, that would be internally inconsistent and lead to contradictions.

 Quote: So, when the international standard of the kilogram appears to be varying in weight, Dr Terry Quinn CBE FRS, Emeritus Director of the BIPM and a recognised world authority in metrology, will happily say "clearly the IPK is changing weight, no matter what the SI definition says in principle." In contrast, GuyBarry, self-confessed mathematician, claims that he's "talking rubbish" because nothing can ever conflict with the perfect world of the SI definition.

Well it wasn't me that came up with the SI definition, it was the BIPM! Dr Quinn is quite entitled to say "relative to other national prototype kilograms, the IPK has lost mass by 0.000005%" (or whatever the figure is). What he's not entitled to say, for as long as the IPK remains the standard for mass, is "the IPK has lost mass by 50 micrograms". The mass of the IPK is 1 kg and remains so until the new definition comes into force. Any other value for the mass of the IPK would lead to absurdity.

 Quote: This is not arithmetic, this is metrology.

Which is based on arithmetic - at least I hope it is.

 Quote: This is not maths, this is physics.

What is based on maths - at least I hope it is!

1304250.  Thu Nov 22, 2018 10:08 am

GuyBarry wrote:
So where's the proof that the speed of light in a vacuum is constant? There isn't one. As this article puts it:

 Quote: It should be noted that Einstein did not actually PROVE the constancy of the speed of light in all frames of reference. Rather, it is an axiom (an underlying assumption) from which he derived the rest of his theory. The axiom can be experimentally verified, but it is not proven in any theoretic sense.

Dammit, I clearly should've gone with the Cambridge Dictionary definition instead. That's a pretty weasely thing to do at this stage, but we've descended into arguing semantics which is a sure sign that a discussion has hit rock bottom, so what the hell:

assumption /əˈsʌmp.ʃən/ [noun]: something that you accept as true without question or proof

Einstein certainly did make the assumption that the speed of light is constant when he formulated his theory of relativity. At which point pretty much the whole of science (quite rightly) said "come off it, Albert, you're talking bollocks."

It was only after experimental data was observed which supported the theory that people started to accept that the German guy with the unwieldy hair might be on to something. Since then, every time the theory has been tested, it's been right. And by "right", I don't mean "in the general ball park" but "bang on the money". Including phenomena which hadn't even been dreamt of when Einstein was alive.

So, if an assumption is something that's accepted without question, then it's not an assumption.

 GuyBarry wrote: And the fact that you admit of the possibility that the speed of light might not be constant indicates that there cannot be a proof that it's constant.

In Physics you can prove something is false, but you can't prove anything is true. All you have is an awful lot of consistent data.

 GuyBarry wrote: But what you can't do is continue with the old definitions and the new ones at the same time.

Most people can. As I explained in the other thread, the SI definitions are a system to define things to absolute precision. The vast majority of science that's done every day doesn't require this level of precision, so it will make precisely fuck all difference to the vast majority of scientists what system is in place, or even if two conflicting systems are in place at the same time.

The theoretically perfect precision of the SI definitions have very little to do with the vast majority of pragmatic science that's done every day.

 GuyBarry wrote: Well it wasn't me that came up with the SI definition, it was the BIPM!

So if anyone should be allowed to say it's bollocks, it should be them!

 GuyBarry wrote: What he's not entitled to say, for as long as the IPK remains the standard for mass, is "the IPK has lost mass by 50 micrograms".

He may not be "entitled" to say that, but that's exactly what he did say. Again, a clear demonstration of the difference between theoretical ideals and a pragmatic scientist admitting that the universe is an imperfect place.

GuyBarry wrote:
 Quote: This is not maths, this is physics.

What is based on maths - at least I hope it is!

Based on, yes. Slavishly adherent to, not so much :)

1304254.  Thu Nov 22, 2018 11:11 am

 dr.bob wrote: assumption /əˈsʌmp.ʃən/ [noun]: something that you accept as true without question or proof

[Edited this section]
If you don't like the word "assumption", substitute "postulate". It is postulated that the speed of light in a vacuum is constant, and this constant value forms the basis of the definition of the metre.

 Quote: In Physics you can prove something is false, but you can't prove anything is true. All you have is an awful lot of consistent data.

Exactly!

 Quote: As I explained in the other thread, the SI definitions are a system to define things to absolute precision. The vast majority of science that's done every day doesn't require this level of precision, so it will make precisely fuck all difference to the vast majority of scientists what system is in place, or even if two conflicting systems are in place at the same time.

Indeed not - but none the less, whatever system is in force must be internally consistent. I'm sure it would have been a hell of a lot easier in practical terms for the BIPM to phase in the new definition of the kilogram over a period of several years, rather than choosing a specific date for it to come into force. But
that would have created an inconsistent system, and the officially recognized system of measurement cannot contain inconsistencies.

That doesn't mean that people won't start unofficially adopting the new definition of measurement before 20 May 2019, of course, and relative to that system the mass of the IPK falls short of 1 kg. But it still officially remains 1 kg right up until that date.

 Quote: He [Dr Quinn] may not be "entitled" to say that, but that's exactly what he did say.

I've re-read the Guardian article now and I must apologize to Dr Terry Quinn. He did not say, as you claimed above, that "measurements over the past century have shown that the international prototype has lost around 50 micrograms". Those are the words of the journalist Alok Jha, who wrote the article. I'm quite sure that a distinguished metrologist such as Dr Quinn wouldn't have made such a straightforward error.

I invite you to apologize to Dr Quinn as well.

[is physics based on maths?]

 Quote: Based on, yes. Slavishly adherent to, not so much :)

I don't know how you can say that. I studied both special relativity and quantum mechanics as part of my undergraduate maths course. The maths was extremely rigorous, and I admit that I struggled with it. It would be impossible to build up such a rigorous body of mathematical theory without accepting the basic principle that 1 kilogram equals 1 kilogram.

Last edited by GuyBarry on Fri Nov 23, 2018 4:16 am; edited 1 time in total

 1304267.  Thu Nov 22, 2018 12:54 pm The Hyperloop facility has a test tunnel one mile long. If it is straight, it provides a vacuum tube 1600 m long. If they extend it to, say, three miles then that gives us approximately 5000 m of vacuum through which to shine a laser beam. Starting with 1,600 m though, place a mirror at one end and a rotating glass prism at the other; essentially the Fizeau–Foucault apparatus. Instead of a cogwheel to interrupt the beam, shine the laser onto the rotating prism, such that it reflects the ray exactly down the tunnel to the mirror only at one point of the rotation. Then measure the angle at which the returning light is reflected when it hits the prism, which will show how far the prism has rotated in the time taken for the light to pass over 3 000 m. When the Fizeau–Foucault experiment was carried out (or was it Michelson?), it was done over a longer distance, but in air, which introduces an error as light is slower in air. I have not heard of vacuum apparatus long enough for running the test until Hyperloop, but do correct me as I have not read the literature. There is a question of how the timing would be affected by the light's undergoing three reflections, and how accurately the rate of spin is measures. It would be interesting to see if there is a measurable difference from the other methods that have been used to measure c since F-F.

 1304271.  Thu Nov 22, 2018 1:07 pm Ah. I just looked on Unreliablepedia and see that everything I said was wrong. Michelson did have a vacuum tube one mile long. He did not have a laser though. If Hyperloop ever build a tube five miles long and straight, with a hard vacuum, then I'd say go for it.

 1304304.  Thu Nov 22, 2018 6:57 pm - although having mentioned the process of reflection as affecting the travel time, in fact the amount of time the light would spend in the medium of the mirror would be infinitesimal and immeasurable and so no more than a theoretical factor.

1304344.  Fri Nov 23, 2018 9:36 am

 GuyBarry wrote: If you don't like the word "assumption", substitute "postulate".

And what if I don't like the word "postulate"? OED says:

postulate /ˈpɒstjʊlət/ [noun]: A thing suggested or assumed as true

And we're back to an assumption again.

 GuyBarry wrote: I'm sure it would have been a hell of a lot easier in practical terms for the BIPM to phase in the new definition of the kilogram over a period of several years, rather than choosing a specific date for it to come into force.

Would it? Why?

 GuyBarry wrote: That doesn't mean that people won't start unofficially adopting the new definition of measurement before 20 May 2019, of course, and relative to that system the mass of the IPK falls short of 1 kg. But it still officially remains 1 kg right up until that date.

"Officially" is the correct word here. Pragmatically, everyone knows that the IPK doesn't really weigh exactly a kilogram, and the only people who would insist it still does are the kind of crazy, anally-retentive people who would insist on following a set of rules despite the fact that they're clearly bollocks*

 GuyBarry wrote: I've re-read the Guardian article now and I must apologize to Dr Terry Quinn. He did not say, as you claimed above, that "measurements over the past century have shown that the international prototype has lost around 50 micrograms". Those are the words of the journalist Alok Jha, who wrote the article.

Yes, you're absolutely right, and I apologise fully and completely to Dr Quinn.

However, here's an article in which Dr Richard Davis (Principal Research Physicist Emeritus, BIPM Consultant, winner of this year's William A. Wildhack Award for outstanding contributions to the field of metrology and measurement science, and the owner of a damn fine moustache!), whilst talking about the IPK, states:

 Quote: we estimate the mass has changed by about 50 parts per billion over 100 years

 Quote: We want to have our reference standard better than that - units need to be defined to at least the precision with which measurements can be made.

*Or "mathematicians" as they're otherwise known ;-)

1304345.  Fri Nov 23, 2018 9:40 am

 Alexander Howard wrote: If Hyperloop ever build a tube five miles long and straight, with a hard vacuum, then I'd say go for it.

No point. Back in 1676, Ole Rømer explained how you can easily measure the speed of light using a minimum of 587 million kilometres of hard vacuum.

1304369.  Fri Nov 23, 2018 10:40 am

dr.bob wrote:
 GuyBarry wrote: If you don't like the word "assumption", substitute "postulate".

And what if I don't like the word "postulate"?

It seems to be good enough for this site: "one of the basic postulates of special relativity".

http://www.einstein-online.info/dictionary/constancy-of-the-speed-of-light.html

But pick your own word if you like - I can't see that it makes any difference. The point is that it's an established principle that underpins a scientific theory, and which is generally accepted as true. What word would you prefer?

Quote:
 GuyBarry wrote: I'm sure it would have been a hell of a lot easier in practical terms for the BIPM to phase in the new definition of the kilogram over a period of several years, rather than choosing a specific date for it to come into force.

Would it? Why?

Well I don't suppose everyone working in every laboratory across the world is going to be able to re-calibrate their equipment on the same day.

 Quote: Pragmatically, everyone knows that the IPK doesn't really weigh exactly a kilogram, and the only people who would insist it still does are the kind of crazy, anally-retentive people who would insist on following a set of rules despite the fact that they're clearly bollocks

So why didn't the BIPM abandon the link between the IPK and the kilogram twenty years ago when it was suspected that the IPK was losing mass? Why has it taken them all this time to bring in a new standard? Despite apparent discrepancies in the mass of the IPK, no one has been able until now to agree on a new standard that was more accurate than the IPK. Here's Wikipedia:

 Wikipedia wrote: After the International Prototype Kilogram had been found to vary in mass over time relative to its reproductions, the International Committee for Weights and Measures (CIPM) recommended in 2005 that the kilogram be redefined in terms of a fundamental constant of nature. At its 2011 meeting, the CGPM agreed in principle that the kilogram should be redefined in terms of the Planck constant, h. The decision was originally deferred until 2014; in 2014 it was deferred again until the next meeting. CIPM has proposed revised definitions of the SI base units, for consideration at the 26th CGPM. The formal vote, which took place on 16 November 2018, approved the change, with the new definitions coming into force on 20 May 2019.

As far as I can gather, it was the development of the Kibble balance that allowed scientists to indirectly measure mass with a sufficiently high precision. Flawed as it may have been, the IPK was actually the best standard for mass available until the Kibble balance was refined sufficiently. It wouldn't have lasted since 1889 otherwise.

dr.bob wrote:
However, here's an article in which Dr Richard Davis (Principal Research Physicist Emeritus, BIPM Consultant, winner of this year's William A. Wildhack Award for outstanding contributions to the field of metrology and measurement science, and the owner of a damn fine moustache!), whilst talking about the IPK, states:

 Quote: we estimate the mass has changed by about 50 parts per billion over 100 years

That's fine. But it has not changed by "50 micrograms". If it's lost 50 ppb relative to some other reference mass, then, by the definition of the kilogram, that other reference mass has increased by around 50 micrograms. There's no way of genuinely knowing whether the mass of the IPK has gone down or the other mass has gone up. All the evidence may point to the theory that the mass of the IPK has gone down; but, until an independent standard has been agreed, it can't be stated for certain.

 Quote: We want to have our reference standard better than that - units need to be defined to at least the precision with which measurements can be made.

Sure. I am not proposing that the IPK should be retained as the standard for mass now that a better one has been found!

 1304379.  Fri Nov 23, 2018 12:09 pm William A Wildhack -- what a superb name!

1304426.  Fri Nov 23, 2018 2:28 pm

 dr.bob wrote: Back in 1676, Ole Rømer explained how you can easily measure the speed of light using a minimum of 587 million kilometres of hard vacuum.

Hang on, I've got some old gas pipe and a bike pump, leave it with me and I'll get back to you!

1305034.  Thu Nov 29, 2018 6:41 am

 GuyBarry wrote: But pick your own word if you like - I can't see that it makes any difference.

I can't be arsed to argue semantics as it'd make a dull thread even duller. All I want to say about this is that the reason why the constancy of the speed of light has been accepted as a basis for relativity is because of the vast wealth of corroborating data. It doesn't matter how elegant a theory relativity is, or how much people like it, without the corroborating data to back it up, it wouldn't be given the time of day.

 GuyBarry wrote: Well I don't suppose everyone working in every laboratory across the world is going to be able to re-calibrate their equipment on the same day.

Is this what you genuinely think will happen? If it is, then it certainly explains why you're getting quite so het up about the definitions.

I'm pretty sure I've explained previously, both on this thread and the last one, that that's simply not how science is done in practise.

 GuyBarry wrote: So why didn't the BIPM abandon the link between the IPK and the kilogram twenty years ago when it was suspected that the IPK was losing mass?

Because they didn't have a better alternative. They knew the IPK was shit, but they hadn't yet come up with something that was less shit.

 GuyBarry wrote: Why has it taken them all this time to bring in a new standard?

Because it's not an easy thing to define in absolute terms and, as I explained above, they were waiting for the corroborating data that suggests that the new definition is a good way to go.

 GuyBarry wrote: There's no way of genuinely knowing whether the mass of the IPK has gone down or the other mass has gone up. All the evidence may point to the theory that the mass of the IPK has gone down; but, until an independent standard has been agreed, it can't be stated for certain.

I don't know how many times I can say "yes, but everyone already realised that the official definition was bollocks" in different ways. It may not have been possible to state for certain that the IPK lost exactly 50 micrograms, but it was possible to state for certain that, as an absolute definition of mass, it was a bit shit.

Page 2 of 3
Goto page Previous  1, 2, 3  Next

All times are GMT - 5 Hours

Display posts from previous:

Forum tools
User tools