One thing that I have noticed over years spent on the Internet is that there isa seemingly large, and very vocal, group of people who spend a lot of time andenergy getting worked up about measurement systems. And, almost universally,their distaste is targeted in one direction: imperial units. Metric good,imperial bad, is the general refrain.
I recently spent a bored evening paging through severalanti-imperial/pro-metric videos on YouTube and, in spite seeing a lot of a lotof emotions and non-sequitur arguments, I didn’t find much in the way ofactually valid and meaningful argumentation. So I wanted to take a few minutesand write out responses to some of the common arguments (or, in many cases,non-arguments) presented in these videos, from the perspective of someone who,horror of horrors, is rather fond of the imperial system.
First things first, we need to establish what we mean by “imperial” and also by“metric” units. The truth of the matter is that these terms mean different thingsin different contexts, and in different time periods, and it is important to beprecise with our language. For example, one interpretation of “imperial” mightbe the English Engineering system with its truly awful pound-mass (lbm) unit formeasuring mass, necessitating an extra conversion factor within equationsbecause its definition is inconsistent with the rest of the system. But thatis, of course, not a system I’m going to be defending.
For the purposes of this post, when I use the phrase “imperial system”,“imperial units”, etc., I am actually talking about the more modern BritishGravitational System, with its base units of foot-slug-second. And when I talkabout metric, I will be referring to the meter-kilogram-second system, alsoknown as SI units.
Neither of those systems have independent volume measurement, using cubic feetand cubic meters as the base volume unit respectively, so I’ll also talk aboutliters and gallons, and their associated units.
Now that we have defined the terms that we are talking about, let’s take a lookat some common arguments that I’ve seen crop up within these videos. I’ll do mybest to represent them accurately and in good faith. We’ll start with themost obvious one.
Metric Unit Conversions are Easier
One very common argument used in favor of the metric system and against theimperial system is that unit conversions within the system are easier. Metricconversions are done in a fully standardized system, only needing a shifting ofthe decimal point. Whereas imperial conversions are much more convoluted. Aclassic example is that $1 \text{ km} = 1000 \text{ m}$, whereas $1 \text{ mi}= 5280 \text{ ft}$.
This is a point that is absolutely true. Performing a conversion fromcentimeters to kilometers is far easier than from inches to furlongs. You can’treally argue otherwise. However, I would argue that this point is pretty muchirrelevant for most practical situations. The fact of the matter is that unitconversions fulfill very different purposes in metric and imperial, and so tocompare the relative difficulty of conversion between two metric units, and twoimperial units, isn’t particularly valid. Metric unit conversions are usedprimarily to condense the representation of numbers, and imperial unitconversions are mainly used to ease measurement or calculations.
In the case of metric, a unit conversion is performed by shifting the decimalpoint around, which changes a prefix on the unit in question. For example,there are 100 centimeters in 1 meter. So 10 meters is equal to 1,000centimeters. This is a very straightforward and easy process–but, at leastin my opinion, it’s largely vacuous.
The reason why I say this is that, effectively, all that this scalingup-and-down of the units accomplishes is reducing the number of insignificantdigits that you need to represent with a 0. If you have 1200 meters, you canavoid writing two of those zeroes by instead writing 1.2 kilometers. But, inpractice, that’s all that you accomplish.
This is made even less significant by how these unit conversions are actuallyused in practice. Without going too deeply into the weeds of dimensionalanalysis, the equations used in science and engineering only work out cleanlywhen the units associated with all of the physical quantities used as inputsto the equation align. As a simple example, in Newton’s 2nd Law of Motion,$$\vec{F} = m\vec{a}$$The force ($\vec{F}$) is conventionally expressed in Newtons. But this onlyworks if mass is in kilograms, and acceleration is in meters per square second.If you were to use a mass in grams, the result would only be in Newtons if youalso converted the acceleration to be in millimeters per square second. All ofthe quantities need to be scaled up or down the same amount–otherwise you’llget a result in a non-standard set of units for force. This is actually why theCGS system defines its own unit of force, the dyne, for using centimeters andgrams together in this equation, and others.
In practice, what this means is that, when working in the SI system to do scienceor engineering, you rarely use kilometers or milligrams or centimeters.It causes too much of a hassle making sure that all of your units areconsistent with each other. Most people simply convert everything into metersand kilograms, and then go from there. And then it is far simpler just to leavethe resulting quantities in the base units as well, because scientific notationhandles perfectly well the compression of insignificant zeroes, withoutrequiring the next person to need to undo your conversions before using theresult in another calculation.
Likewise, it is very rare in real life to convert from feet or yards into mileswhen using the imperial system. It isn’t as though people actually use these assubdivisions of one another in most cases. You’d never see someone reporta distance as “5 miles, 1571 yards, 2 feet, and 9 inches”. You will see peopleuse feet and inches alone like this (and sometimes yards), but never will milesbe thrown into the mix. And the conversions for inches, feet, and yards are areeasier than the strawman example of feet and miles that is commonly bandiedabout. They’re still more complex than simply shifting the decimal, like inmetric, but for the additional complexity, these conversions actually aid inperforming calculations with these units. For example, if you wanted to subdividea foot into 3 equal parts, it’s much nicer to think in terms of 4 inches than interms of .33333333333333333333… feet.
So yes: it is true that unit conversions are far easier in metric than they arein imperial. But the commonly provided examples of difficult imperialconversions are, in practice, almost never used, and the commonly usedconversions in imperial are relatively simple (though, admittedly, not assimple as metric).
The Imperial System is a Mess
I’m not going to write at length on this one because it seems to me to belargely a non-sequitur. You’ll frequently see this image bandied about asthough it proves something,
This image shows the relations between many different units of length used inthe imperial system, with their conversion factors. And it’s true, as renderedin that figure the system seems very complex compared to metric, which doesn’teven warrant a diagram as the only unit of length relevant there is the meter.
The trouble is that this image is also incredibly misleading. The imperialsystem is a lot older than the metric system, and has picked up a lot ofadditional baggage of the years. Most of the units shown on this diagram areeither (a) not in use any more, or (b) used in very specific contexts.
Unless you’re doing surveying work or typography, it is likely that the onlyunits that really matter on this chart are inches, feet, yards, and miles.
That said, there are examples of imperial units where this critique is valid.One commonly noted example is the definition of a “barrel”, of which there areseveral versions used in different contexts. Similarly there is the short andthe long ton. But, again, these distinctions arise in very specialized areas.And the barrel situation can be fixed via standardization within the imperialsystem, without a conversion to metric.
Similar odd units crop up in metric too, like shakes, angstroms, light-years,light-nanoseconds, parsecs, electron-volts, foes, ergs, etc. One might arguethat these are obscure, or relevant to very particular scientific domains. Towhich I would reply that the same can be said of links, cubits, and skeins.
The Mars Climate Orbiter, Air Canada Fueling Incident, and other Such Examples
Another common argument used against imperial are accidents that were causeddue to either misconversions, or miscommunications in areas where both metricand imperial units were used concurrently. These are not arguments specificallyagainst imperial. The accidents that occurred were not due to the imperialunits themselves, but rather due to environments in which two unit systems wereused together. These examples demonstrate that using only a single unitsystem at a time in a given context is a good idea, but don’t provide anyspecific condemnation of imperial. The same could have happened at theinterface between two metric systems (Like SI and CGS).
Scientific Literacy
One particularly interesting argument that I saw was brought up by Kurtis Bauteat around 4:30 in this video. Heargues that science is fundamentally about taking measurements–which isabsolutely true, and that being able to accurately take a measurement isimportant to science and scientific literacy, which is also very true. He thengoes on to say that the use of imperial units prevents people from knowing“what a meter is” and implies that this prevents them from being able to doscience.
Okay, so to be fair I did pick a pretty over-the-top variation of this argumentto lay out. Generally speaking, science is done using metric, and so in orderto participate in science, one must learn metric. This is true. However, I don’tthink that the use of imperial units within the US significantly hurts metricl*teracy.
For one thing, metric is taught in schools in the US. I certainly learnedit, and I’m sure most other Americans did as well. In fact, for a few yearsI managed a general physical sciences course at a university that I workedat–the sort of base-level exposure course that every first-year studentwas required to take–and I was consistently surprised by how almost everystudent was much more comfortable with metric than with imperial. I think thatit has to do with the fact that the imperial system relies heavily on fractions,rather than simple decimal shifting, which seems to be a trouble-spot for a lotof students.
As a side note here, I sometimes feel that the metricization of science in highschool goes perhaps a little too far. It is not uncommon when judging highschool science fairs to see very strange quantities in either experimentalprocedures, or in results sections. For example, one might see a methodssections including “56.7 grams of baking soda were added to 236.59 millilitersof water” as one of the steps. Why such strange numbers? Because the student inquestion was doing an experiment in their kitchen, and measured out 2 oz. ofbaking soda to add to 1 cup of water. But because “SCIENCE MUST BE METRIC!!!!”,instead of reporting the precise measurements that they took on the instrumentsthat they had available, they lose precision and do silly and unnecessaryconversions for no good reason. To accurately report the measurements that theytook in the form that they took them would result in losing points in judging,because metric is the only acceptable unit system in science.
Convenient “rough” Estimates between Dimensions
Another advantage of the metric system is the convenient conversion between thestandard SI volume units, and the non-standard but commonly used volume unit ofliters. Specifically, one cubic centimeter is equal to one milliliter. Andfurther, 1 milliliter is roughly equal to 1 gram of water. Yes–this is arough and somewhat problematic conversion, but it gets brought up a lot.
Taking an example from this video,then, can we easily calculate how much water we can haul in our truck,with a max payload of 1,000 kilograms,$$1000 \text{ kg} * \frac{1 \text{ L}}{1 \text{ kg}} = 1000 \text{ L}$$
What he doesn’t mention is that a similar conversion exists in imperial, onepint of water weighs about 1 pound. It isn’t quite as precise as the metriccase, but is close enough for practical use (it’s about 4% off). So to do thesame problem in imperial units, (using 1 ton instead of 1000 kilograms)$$2000 \text{ lbs} * \frac{1 \text{ pt}}{1 \text{ lb}} = 2000 \text{ pt}$$Now if we wanted the result in gallons, we’d need to divide by 8, whichI admit isn’t quite as convenient as shifting a decimal place. But itstill isn’t too bad: 250 gallons. The correct answer is 260–which isslightly under 4% higher than the estimate. Not too bad for a stupid,horrible unit system.
In either case–this is very much in the “rough” calculation territory inprecise situations, because the density of water is actually a function oftemperature and pressure, and so such calculations will always have a margin oferror unless these other variables are also accounted for. At which pointa thermodynamic table must be consulted, and any illusion of “speedycalculation” goes out the window in both cases.
Metric Units are Fundamental
These days, the metric units have all been defined in terms of universalconstants. Thus, these units are fixed to the fundamental nature of theuniverse in some way.
Of course, these definitions are retroactive. The units had already existed,and were simply tied back to some fundamental constant with a random conversionfactor applied to make sure that the defined value matched up with the originalone. And, of course, imperial units are defined in terms of metric ones now,and so have just as strong an argument to “fundamentality” as the metric onesdo. If you’re going to define a meter with the constant $\frac{1}{299792458}$relative to a fundamental constant, and a foot with the constant$\frac{1}{913767411984}$ relative to that same fundamental constant, can youreally claim one is “more fundamental” than the other? Those are both prettyugly numbers.
The Imperial System uses Pound for both Mass and Weight
It used to. It doesn’t anymore. The British Gravitational System does away withpound-mass and has the slug as a unit of mass instead. No ugly conversionsneeded. However, the metric system, in general use, uses the kilogram for bothmass and weight. If anything, the imperial system is more accurate toreality here, measuring weight with a unit of force.
After all this, I do want to say that I do like metric units and use them quiteregularly. But I also like imperial units, and use them regularly too. I’llconcede that there are definitely situations where metric units are better thanimperial (drill, tap, and screw sizing come to mind…). But I get consistentlyannoyed by the metric elitists bandying about bad arguments for their goodsystem, and so I wanted to address some of those arguments here. Imperial unitsare not automatically dumb, imprecise, or non-scientific. And you can doperfectly good work no matter which unit system you choose.
