Talk:International System of Units/Archives/11/2019

Insistence on citation

@DeFacto: I'm mystified by your your insistence on a citation needed tag for a statement that is linked to an article on the topic that is comprehensive and richly cited. This is just a general example being given here in a nonessential section, not a detailed claim about a precise number. Are you challenging the veracity of the statement "The metre is close to the length of a pendulum that has a period of 2 seconds"? Or do you feel that every claim made in WP should have citations, regardless of whether they have links to well-cited articles that give notable references for anyone who might be interested in verifying the claim? Shouldn't we rather focus on making the statement more relatable (at least to those who might have seen one of these) by, for example, mentioning a grandfather clock which typically has a seconds pendulum? —Quondum 19:25, 20 November 2019 (UTC)

@Quondum: firstly, your first link above is, I guess, not the one you intended. Secondly, I'm concerned about the low quality of sourcing in this article in general, and when this substandard one was removed and replaced with a "citation needed" flag, and then that flag was removed without any explanation at all, I thought it should be restored. I think that each article should be self-sufficient, in terms of sourcing, and we should not expect a reader to hunt them down in other articles. -- DeFacto (talk). 22:18, 20 November 2019 (UTC)

Thanks for pointing out the link error; I've corrected it. I did not find your first of the two reverts strange (where the reverted edit was without an edit summary), but that does not apply with the second revert. Whether sourcing is an issue in general in this article is not really pertinent here. My befuddlement is centred around the following observation: if the bar for needing a source is set such that that statement needs a source, then effectively every single statement in every article needs a source. It does not seem appropriate to leave {{cn}} tags on any but those that may seem implausible or that are being challenged. The guideline for sourcing states "The policy on sourcing is Wikipedia:Verifiability, which requires inline citations for any material challenged or likely to be challenged [...]" The documentation of {{cn}} says "{{Citation needed}} is a template used to identify claims in articles, particularly if questionable, that need a citation to a reliable source" [my emphasis]. You claimed in your edit summary of the revert that I linked above that "this does need an RS", but you have failed to point out why this particular statement might. —Quondum 00:33, 21 November 2019 (UTC)

This is a subjective and imprecise claim ("close to the length") and inquisitive readers may wonder just how close. Clicking-through to the linked article is not the answer. It is a poor quality and rambling article and a source for the length doesn't jump out at you. WP:VER is also very clear that "Any material lacking a reliable source directly supporting it may be removed and should not be restored without an inline citation to a reliable source." -- DeFacto (talk). 07:31, 21 November 2019 (UTC)

The relationship between the length of a pendulum and its period is taught to sixteen-year olds at school. Or at least it was taught at my school when I was 16. I think stuff that we learn at school can be considered both widely known AND widely accepted, and therefore uncontroversial. A link to the pendulum article is sufficient IMO. Dondervogel 2 (talk) 08:25, 21 November 2019 (UTC)

@Dondervogel 2: so what percentage of Americans would you imagine know that the length of a pendulum that has a period of 2 seconds is close to 1 metre? -- DeFacto (talk). 20:06, 21 November 2019 (UTC)

All of the ones who read this article carefully enough. Dondervogel 2 (talk) 22:30, 21 November 2019 (UTC)

And how many do you think knew it before reading the article? -- DeFacto (talk). 22:35, 21 November 2019 (UTC)

DeFacto, there are two major problems with your argument:

• You fail to acknowledge, despite my highlighting it, that WP:RS and WP:V both emphasize that they apply to "material whose verifiability is challenged or likely to be challenged". You are apparently carefully skirting around this by failing to acknowledge my questions, and by ignoring the paragraph immediately preceding the one you chose to quote, to the point that a charge of not listening may be warranted.
• You appear to be confusing the readers' interest or lack of knowledge ("inquisitive readers may wonder just how close") with a requirement for verifiability, making your argument unrelated to policy and guidelines.

—Quondum 02:46, 22 November 2019 (UTC)

@Quondum: how do readers know that sentence isn't just made up? How can they know it comes from a reliable source? The easiest way to convince them is to supply a reliable source supporting it. I cannot see why the reluctance to either supply one, leave the request in place, or perhaps remove the assertion altogether. I don't understand either of your points:

• the first; this has been challenged already, so is moot.
• the second; you seem to be trying to obfuscate the obvious need for a source. We have already had a challenge, so we know a source is needed (see WP:BURDEN). But, even if we hadn't, we cannot know the educational background of readers, so cannot assume that they all have a sound understanding of physics (see WP:AUDIENCE). Thus this science-related assertion needs to be sourced per WP:UNSOURCED.

-- DeFacto (talk). 17:20, 22 November 2019 (UTC)

Revolutions and radians

@Quondum: I am definitely not proposing to equate ω (omega, the pulsation, or angular speed, SI unit radians per second) and ν (nu, the frequency, SI unit hertz, or revolutions [or cycles if you prefer] per second). What I have is ω = 2πν, precisely because one revolution is equal to 2π radians, and both revolutions and radians are dimensionless units (if they were different units, as mass and weight are, I couldn't equate the one to the other multiplied by some pure number); so if ν is 100 Hz, ω will equal 628.32 (± 0.01) rad/s. Or if an engine is running at 3000 rpm, its axle is rotating at (3000 * 2π / 60) = 314.16 (± 0.01) rad/s. What I had been writing on the article page (and you removed) is equivalent to saying that a frequency of 3000 rpm is identical (via a change of viewpoint) to an angular speed of 3000 turns per minute, which is equal (via a change of units) to an angular speed of 100π radians per second. (One turn, or one "full circle" turn, is an angle of 360 degrees or 400 grads or 2π radians etc.). (And if what I'm saying here seems elementary to you, sorry, it's professional deformation: I'm a retired secondary school math and physics teacher, and the above is exactly the kind of thing it used to be my job to teach.) — Tonymec (talk) 23:26, 21 November 2019 (UTC)

P.S. More precise, maybe, than ω = 2πν would be saying that ω = ν * 1 turn, i.e., if angles are expressed in turns then ω_tr = ν; if they are expressed in radians, then ω = 2πν; if they are expressed in degrees and decimals then ω_deg = 360° * ν etc. — Tonymec (talk) 23:39, 21 November 2019 (UTC)

Although the edit contained a small error (missing brackets around 60 s) I agree with Tonymec's reasoning. Dondervogel 2 (talk) 23:44, 21 November 2019 (UTC)

P.P.S. The pulsation comes in useful as a shortcut for 2πν when talking of phenomena which are not rotating but oscillating: then one writes y = y_max sin ωt, where trigonometric ratios are implicitly expressed from angles in radians, as a shortcut for y = y_max sin 2π(t/T) where angles are in radians and T (the period) is the inverse of ν (the frequency) (and t is of course the time). — Tonymec (talk) 00:22, 22 November 2019 (UTC)

The statement in question is "2π/60 rad/s = 1/60 Hz". If we put Hz = s⁻¹ and rad = m/m (both stated in the 9th SI Brochure), then upon substitution 2π/60 rad/s = 1/60 Hz implies 2π = 1. —Quondum 02:05, 22 November 2019 (UTC)

That's what you get when you assume that different units for a single kind of variable (here: angle) are necessarily the same. Angles can be measured in turns, radians, degrees or grads but it doesn't follow that 1 turn = 1 rad = 1° = 1 gr. Hertz are turns per second, or have dimension s^-1 since turns are dimensionless. Radians per second also have dimension s^-1 but they aren't hertz, they are 6.28… times smaller. — Tonymec (talk) 04:14, 22 November 2019 (UTC)

P.S. Radians are metres per metre but they aren't just any quotient of two lengths: radians are metres of circumference along a circle per metre of radius of the same circle. You cannot divide any two lengths and say you got radians, it doesn't work that way. For instance, metres of length by metres of width (of a rectangle) are metres per metre, but they aren't radians: radians are a unit of angle, and length per width is a certain quantity which I shall henceforth call aspect. For instance, a square (length = width) has an aspect of 1. All European standardized paper sheets (A0, A1, A2, etc.) have the same aspect: their length is √2 times their width; but you couldn't find √2 radians anywhere in them. About physical quantities, most have a definite dimension; but a few, like the turn, the radian and the decibel, are dimensionless; which doesn't mean they are equal: one turn is equal to 2π radians, but by no stretch of mind (that I can think of) could either of them be converted into some number of decibels. Where it gets hairy is that some quantities are usually expressed in terms of turns, but could in some applications be converted to radians (as with a frequency in cycles [i.e. turns] per second and the corresponding pulsation in radians per second, which are just two different ways of looking at a single thing, but stand in a ratio of 1:2π in relation to each other); and the opposite also happens (as the angular speed of a transmission axle, which in SI units would be some number of radians per second, is usually expressed as a number of revolutions [i.e. turns] per minute, which are not just 60 times the SI value of the angular speed, but 60/2π). Radians and turns have the same dimensionality of 1 (i.e. "pure numbers") but one radian is not one turn. I hope I made myself understood. — Tonymec (talk) 05:23, 22 November 2019 (UTC)

You don't get that the equation 2π = 1 is incorrect, no matter how you rationalize it? When your reasoning leads to a contradiction, it is a fallacy. Your argument seems to come down to "2π = 1, because each side is in different units". The SI does not permit equations that are nonsense; it has (regrettably, IMO) chosen to make both angle and count (or cycles) truly dimensionless. The SI resolves this by saying that the quantity that we measure in cycles and the quantity that we measure in radians are as distinct as the circumference and the radius of a circle. While they are quantities of the same kind, they are unequal; that is to say, 2π rad ≠ 1 cycle (in the SI). —Quondum 14:17, 22 November 2019 (UTC)

• This revert was correct because 1 rad is not equal to 1/(2π). (If you think it is, think about whether 1 mol is 6.02214076×10²³; then whether 1 s is equal to 1/9192631770; then whether 1 m is equal to 1/299792458 s; then about the nature of units in general.)

However, the example is ill-chosen because "rotation per minute" can be either a frequency or an angular frequency unit (whose values coincide, but as the discussion above shows, can be confusing), so I removed it. Tigraan^{Click here to contact me} 14:42, 22 November 2019 (UTC)