The flip side of SI

Charlie P, a regular if occasionally critical contributor to MV, promised he would on his return from holiday give his thoughts on “the flip side of the SI as far as the average man in the street is concerned”. These now appear below. They are of course not endorsed by the Editors of MV, but may provide food for thought.

Lack of empathy with non-scientist human users

* Not optimised to accommodate human frailties and weaknesses.
* Not accessible to all strata of society.
* Commonly used units are not based on a human-friendly scale or with reference to readily and universally available cue objects.
* The units do not lend themselves to accurate approximation without the use of measuring tools.
* The strict base-10 system does not facilitate divisibility by multiple factors (whereas base-12 does).
* The strict base-10 system does not facilitate mental arithmetic, which is easier with fractions than with decimals.
* The size of the metre derives from a ten-millionth of the distance from the equator to the north pole through Paris! Hardly a human scale.
* The size of the gram derives from the weight of one-millionth of a cubic metre of water! Again, not a human scale.
* The rules of writing metric units are so complex and so obtuse that very few people outside of academia are competent in using them. Examples of incorrect usage can be found on government websites, in newspapers, on food packaging, on road signs, on measuring instruments, and even on the UKMA website.

Irregular and inconsistent

* The non-decimal traditional second, based on 1/86,400th of a day, is used as the base unit of time. For some reason(!) the redefinition of the second to be 1 day/100,000 was rejected.
* The base unit of mass (kg) already has a prefix, so breaks the prefix use norms, and adds a level of complexity to the teaching and understanding of the system.
* The symbol for the litre (L or l) is not case sensitive, unlike all the other symbols.
* Land area is measured in hectares (ha), an irregular use of the “hecto” prefix on the now non-existant “are” unit.
* Multiple names for exactly the same unit exist. Examples: 1 sievert = 1 gray, 1 tonne = 1 Gg, 1 litre = 1 dm[^3] (one cubic decimetre), 1 hectare = 1 hm[^2] (one square hectometre).
* The most common speed unit does not have a power-of-10 factor with the SI unit of speed: 90 km/h = 25 m/s.
* Time units (day, hour, minute, second ) and angle units (degree, minute, second) are not decimally power-of-ten factored.

Difficult to use

* Long and complex multi-syllable unit names, for which abbreviations are prohibited. Examples: “kilometres per hour”, “watt per square metre steradian” (unit of radiance).
* Case-sensitive unit symbols can make stencilling/Dymoing impossible. Try spray-stencilling “25 kg” on a sack of potatoes using a lowercase-free stencil.
* Duplicated symbols. Examples: d = deci, d = day. h = hecto, h = hour. m = milli, m = metre. T = tera, T = tesla.
* Reliance on the use of non-roman characters (mu, omega, degrees) and superscript/subscript text (for squared and cubed) can make typing/stencilling,Dymoing impossible. Try making a set of Dymo labels to stick on a rack of resistor storage drawers.
* Case sensitivity of symbols, coupled with the use of non-roman characters, introduces a massive risk of writing and transcription errors. Example: a drug dose of 1 mg (milligram) is one-thousand times larger than a drug dose of 1 [mu]g (microgram).

Too much room for ambiguity and confusion

* The symbol “m” is used for both metre and milli.
* The symbol “T” is used for both tesla and tera.
* The litre has two symbols: “L” and “l”.
* “Mm” “mm”, “da” “dA”, “Pa” “PA” “pA”.
* “Mil” can mean millimetre (mm), millilitre (ml or mL) or milligram ([mu]g).
* Milligram (mg) and microgram ([mu]g) are readily confused.
* Millimetre (mm) and micrometre ([mu]g) are readily confused.
* The SI is defined in French, and its English translation uses unidiomatic terms.
* The writing of metric measures is so fraught with the danger of ambiguity that a whole section of the “SI Brochure” is dedicated to trying to codify it.

Too bureaucratic

* Units are strictly categorised as: “SI base units”, “SI derived units”, “Non-SI units accepted for use with the SI” and “Other non-SI units not recommended for use”. To use SI correctly, users need to know which unit is in which category, and which category is permitted for their particular use.
* The use of the most natural and commonly used abbreviation for “kilometres per hour” (kph) is prohibited by the SI.

27 thoughts on “The flip side of SI”

  1. Strange, but why are we only hearing these shortcomings from the English and the Americans and no one else? Shouldn’t the Germans, the Chinese and other nations with strong debt-free economies be making the same complaints? Yet, from the rest of the world there is total silence and total acceptance of SI.

    So is it SI that is the problem or the people that have refused to adopt it and are now suffering and need an excuse to keep them from having a panic attack?


  2. Even if all these criticisms are taken at face value, SI is still better than the alternatives.

    SI it is accessible to all strata of society, and like decimal currency, is easier for all to compute with than the alternatives.

    Commonly used units work well enough. That’s why they have been so widely adopted around the world. The fact that a metre is one ten millionth of whatever doesn’t matter a bit when you are using a tape measure, but is helpful when you are considering the whole world.

    The “unidiomatic” words for SI become idiomatic after you use them enough.

    Criticising SI both because it’s too regular (base 10) and because it’s not regular enough (e.g., time) seems a bit inconsistent.

    One practical thing about SI is that it’s now used throughout the world. The advantage of almost everyone using the same measures overwhelms the defects and disadvantages that have been noted.


  3. So Charlie P, are you trying to tell us that 95% of the world’s population, and every one of its countries (with one major and two very minor exceptions), have got it all wrong?

    YOU may have trouble dealing with SI, but obviously nearly 7 billion (or 7 000 million if you prefer) do not. Your thoughts will not change a thing. So why bother? What you have written is simply a waste of space (involving old arguments and red herrings that have long been discounted), and a waste of my time reading it.


  4. Wow. Too much to address point by point. A few thoughts:

    A yard is perfect and “human-sized” but a meter is clumsy and awkward, despite a difference of less than 10%? A liter lies between a US quart and a UK quart; both US and UK quarts are human-friendly (the US quart is a former British quart, so hard to criticize) but the liter is clumsy and unfriendly. A “hand” is perfect for measuring horses, but a decimeter (less than 2% different) is clumsy and awkward? A pound is perfect for weighing things out but 2.2 lb (kilogram) is unbelievable clumsy? A tonne is very close to the UK long ton, and fairly close to the US short ton, but clumsy?

    Show me the twelves:
    Sure, there are 12 inches in a foot and 5280 ft is divisible by 12, but twelfths of a mile are rarely used, as are twelfths of an inch. However, the pound is divided into 16 ounces or 7000 grains, where’s the 12 in either? The next unit up is the 14 lb stone. Neither the US (2000 lb) nor UK (2240 lb) ton is divisible by 12 . Your gallon has 160 fl oz, ours 128; where’s the 12? An acre (43560 ft²) is divisible once by 12, but twice by 11; what’s up with that? There are not many factors of 12 in Imperial or Customary; the claim that they are duodecimal in nature is not supposed by evidence. They are a hodge-podge of random factors, a few of which happen to be 3 and 4 (product is 12), but also 5, 7, 11. Well, at least there are no 13’s or higher primes that I can find.

    Non-SI units
    The SI defines a few terms as accepted but not really SI. There are concessions to traditional for the early agricultural and commercial practices of metric, but you praise Imperial’s adherence to those traditions while criticizing the SI (liter, tonne, I admit the hectare is more debatable). There are some other concessions in units of time and angular measure, but they have to be converted for computation. You criticize the SI as bureaucratic, but when it makes concessions to prevailing practice, you criticize it as not bureaucratic enough. Basically, “the SI is bad, I’ll come up with reasons.”

    There are no real differences in SI base and derived units; in calculations, there may be opportunities for unit cancellation and simplification in the derived units. The non-SI units accepted for use with the SI may be convenient in commerce but add a layer of complexity (rather minor) to engineering calculations. “Not recommended for use” means just that; not too complex.

    The tonne is 1 Mg, not 1 Gg
    Sievert ? Gray: The units appear to be the same, but the gray is based on the actual energy absorbed. The sievert is an equivalent dose, the actual energy is multiplied by a quality factor specific to the type of radiation reflecting damage on human tissue (and varies by part(s) of the body exposed).

    I’ve left many points untouched for others to comment on. With the exception of the kilogram having an awkward name, there are few points I can agree with or find a convincing argument that 95% of the world should seriously consider adopting either Customary or Imperial (if so, which???). Should we really return to the Tower of Babel era in which every country has one (or more) sets of incompatible traditional units? I think not, and I think it is not worth the trouble for the US or UK to fight to maintain theirs.


  5. As Charlie P notes, SI (the International System of Units) is not perfect. It might be worth looking at the origins of the metric system.

    On the eve of the French Revolution (1789), the system of units used in France were a mess. The commission that was appointed to look into the reform of the units of measure realised that counting and systems of measurement should have the same base. From a mathematical point of view, base 12 was preferable to base 10, but it was highly likely that a conversion from a decimal to a duodecimal system of counting would be rejected outright, so they decided to live with the impracticality of using 10 rather than 12 as the basis for the new system.

    The commission proposed dividing the day in a decimal manner, a recommendation that was accepted by the French Revolutionary Government. Since the decimal calendar brought no discernible benefits to the population at large, it was discontinued after a few years.

    CharlieP made a few mistakes in his posting:

    Firstly, the unit names “litre”, “hectare”, “tonne”, “hour”, “minute”, “day” and the angular units “degrees”, “minutes” and “seconds” are not SI units, but are units that outside SI, but which, due to their use worldwide, are catalogued in the SI Brochure (and also in ISO 80000) to ensure that people who use them, do so consistently.

    Secondly, he stated that “1 sievert = 1 gray”. They might have the same dimensions and they might be closely related, but they are subtly different – the amount of radiation to which one is exposed is measured in sieverts and the amount of radiation that one absorbs is measured in grays. In the case of X-rays, the strength of the X-Ray beam is measured in Sieverts, but the energy absorbed by the bones and tissue (which is measured in grays) is different, hence the X-Ray pattern that shows up bone fractures etc.

    Finally, may I remind CharlieP that the “I” in “SI” stands for “international” and that the symbols used in SI are designed to be used internationally, not just by the “Man in the Clapham Omnibus”, but also by the “man (or is that person) on the Paris Metro”, the “banana stall-holder in India” and the “captain of the dhow in Zanzibar”. One cannot expect everything to be tuned for the benefit of the British, especially when Her Majesty’s Government pooh-poohed the Convention of the Metre in 1875 and only grudgingly signed it a few years later when they needed access to the prototype metre in order to cross-check whether or not the prototype yard was shrinking (which it was). Also, in the nineteenth century it was the norm that the French language version of diplomatic treaties (which includes the Convention of the Metre) should be the authoritative version of the treaty which is why the authoritative version of the SI brochure is the French version.


  6. -The strict base-10 system does not facilitate divisibility by multiple factors (whereas base-12 does)-
    This implies that the Imperial system is based on 12. However, only 12 inches to the foot and 12 troy ounces to the troy pound. That is all. The metric system may be strictly decimal, as it should be, but it allows for the use of modules which give one back divisibility. One example: the 30 cm module which is close to a foot. It can be divided by 2, 3, 4, 5, 6, 10 and 12 (even two factors more than the Imperial foot!) and each factor gives a rational metric outcome. It seems that houses are built to that module. Modules which allowed for divisibility were used from the beginning. However, never should unit names, like metric foot be given to the modules: use the best of two worlds – the decimal system and divisibility. Keep the secret that in fact you are using one of the oldest measuring units in a metric environment to yourself.
    Syllables: I would rather pronounce a few more syllables than having to calculate in British Imperial or US Customary; for instance making up the Englsh invoices of old which involved Lsd money and tons, hundredweights, quarters and pounds.
    The metre is no longer based on the length of the meridian. The definition of the metre did not include the meridian of Paris. That meridian was measured, but the founders of the metric system assumed that the definition was valid for all meridians.
    Last, but certainly not least, Imperial and US Customary units are now defined in terms of metric standards, making them second hand metric systems, like the trash devised by Napoleon in 1812.


  7. The merits of the imperial and metric systems have been examined as far back as 1862 when the Weights and Measures Select Committee produced a Report favouring the adoption of the metric system. Alternatives examined by the Committee included adapting the imperial system to use a common base and keeping the present system. Hardly any witnesses who appeared before the Committee defended the present system. The Committee also concluded that a decimal system was superior to a duodecimal system. Proof of the inadequacies of the imperial system introduced by the 1824 Weights and Measures Act could be found by the use of different systems for different needs by different groups of people. As the 1862 Report put it:

    “The silent influence of usage has baffled the decrees of legislation; and we are still far distant from the uniformity at which we have so often, yet so vainly, aimed. Omitting many specific anomalies, we have no less than ten different systems of Weights and Measures, most of them established by law. Our neighbours, the French, and many other nations, have only one, founded on the mètre, which is a near approximation to the English yard.”

    You can read more about the 1862 Report at this link:

    Click to access met1862.pdf

    The Hodgson Committee also looked at the issue in 1950 and favoured the adoption of the metric system and the abolition of the imperial system. Here is a quote from their report:

    “The metric system is a closely defined and universally recognised system under the guidance of an international body. The imperial system is a conglomeration of units which form a rough whole. In Great Britain there are five different systems of weight and three of capacity.”

    The 1965 Metrication Programme was begun in the UK at the behest of British industry where the economic benefits and trade advantages were clearly recognised. All major Commonwealth countries have completed their transition to the metric system and are not going back to imperial. Of all the developed countries, only Britain and the United States are still holding out for the continued use of medieval units of measurement.

    While the metric system is not perfect, it is better than the alternatives and the most successful and widely adopted measurement system in world history, now used in every country in the world. Unlike the imperial system, there are no measurement tables to learn. It is vastly superior at measuring extremes at both ends of the scale (e.g. thickness of paper, widths of transistors on computer chips, distances across the universe, etc.) and its multiples and subdivisions cover a vastly bigger range for all types of measurement at both ends of the scale.


  8. Guess we’ll have to wait until there’s a new planet to colonise before all these Luddite measurement anomalies can be rectified.


  9. I did set about answering this item by item. That proved too lengthy.
    In short I came up with “all the usual claptrap, hardly a sensible comment in any of it”.
    Not worthy of further comment.


  10. Some points are good, others not (of course, IMO)… BTW, a future SI could integrate some interesting things – such as systematic and scalable prefixes – from the dozenal SDN/TGM nomenclatures*; while switching the SI to base 12 would probably be impractical, at least for the time being.

    * In particular, the notation that puts the powers of 12 (10 in our case) as subscripts/superscripts to the left of the unit: quite clever, indeed – and no need to constantly invent new abbreviations…


  11. Anyway, at least one thing is really correct, IMHO, in that list: the gram is now a completely “out of scale” unit, also when compared to the metre.

    The metre cannot be changed – it could perhaps have been a little longer, for example like the ancient Roman (double) pace, – but the gram could and should be superseded, as a “virtual” base unit.

    A complete revision and modernisation of the SI would anyway require much more progressive times than today – times which will certainly come, eventually.

    Of course, once the whole world – thus, also the UK and US – has completely adopted the current SI!

    BTW, talking about time, its measurement on the human scale still depends on the Earth’s rotation and revolution, but something could certainly be done also here, at least partially: for example, first of all, adopt the 24-hour clock everywhere; and then we’ll see what can be done, in order to rationalise and simplify the measurement of time…


  12. @Daniel

    Do you deny that the SI has these “shortcomings”?

    Do you still think there is a causal relationship between national debt and level of adoption of the SI?

    Please explain your sentence: “So is it SI that is the problem or the people that have refused to adopt it and are now suffering and need an excuse to keep them from having a panic attack?”. Particularly how, say, the inconsistency in having a prefix on the base unit of mass, could be seen as a problem of “the people” and not of the SI, and what you mean by “are now suffering” and “need an excuse”.


  13. @John Frewen-Lord

    You asked am I “trying to tell us that 95% of the world’s population, and every one of its countries (with one major and two very minor exceptions), have got it all wrong?”

    Not at all, what gave you that impression? I was merely detailing some of the weaknesses of the current SI as far as the average man in the street is concerned.

    Then you said of me: “YOU may have trouble dealing with SI”.

    Again, where did you get that impression? I have no problem dealing with SI. I use little else for work. What I DO have a problem with is tolerating that dogma that the SI is a panacea and that it should be forced into place even where it is unrequired and unwanted.

    Do you have any convincing evidence to support your personal view that “obviously nearly 7 billion (or 7 000 million if you prefer) do not [have trouble dealing with SI]? We have never seen any here.

    Finally you asserted of me that “What you have written is simply a waste of space (involving old arguments and red herrings that have long been discounted)…”

    Can you be more specific about which, in your opinion are old arguments that have long been discounted (and discounted by whom) and which are red herrings that have long been discounted (and discounted by whom).


  14. @Michael Glass

    You say “SI is still better than the alternatives.”

    I agree that it is better than the alternatives in some ways, but not in all ways. We don’t often hear of its weaknesses here, and I was asked to describe where I thought it fell short, so that is what I did.

    Sure some of those weaknesses can be worked-round, others might be more important for some people than others, most are probably tolerable, but they nevertheless exist, and we must not forget that when extolling its virtues.

    By the way, being base-10 doesn’t make SI “too regular”. Totally base-12 or totally base-16 would be similarly regular. No, the SI is irregular in many ways, which can by no means be characterised as a good thing – so deserves criticism.


  15. It would nice to see comments from UK Members of Parliament, Dept for Transport, Met Office and the broadcasters. They do not seem to be interested in progressing UK metrication.


  16. @Rob:

    Politicians in general, and British politicians in particular, have absolutely no interest in anything to do with measuring things, even though measuring things is fundamental to any and every aspect of human existence. Very sad really. The total cost (hidden of course) to the nation in not completing metrication must be tremendous.


  17. I have another comment to make: people only need to know when using SI what they need. The man in the street does not need to know everything. It is also easy to
    learn and use the symbols you need.He condemns SI for being to bureaucratic, but when SI makes concessions to daily use, for instance the time units, he condemns it for being irregular and inconsistent. I know how British Imperial and US Customary work and I find lots of both confusing: miles? what mile do you mean? Ounces, pounds, gallons? Avoirdupois, troy, British gallon, US gallon? Decimalised BI or USC units: utterly absurd. And do not forget the often difficult calculations one has to make or the many conversion factors users of these sets of units need to know.
    I assume that Charlie P. would like to see a ‘new’ SI emerging, now meaning Systeme Imperial, to be adopted by the entire globe!


  18. CharlieP

    I read your latest comment with interest. I think you are making the perfect the enemy of the good. SI is better than the alternatives, and that is a good reason for its adoption worldwide.

    What is clearly worse is where different systems of measurement are used side by side. Things like having ordinary trousers measured in centimetres and jeans measured in inches, or having some rural land described in acres while other land is described in hectares. Then there is the anomalous situation where plane heights are measured in feet but the hills and mountains that they fly over are measured in metres.

    As far back as 1215, the Magna Carta said there should be one measure. Therefore, because SI is better than the alternatives and is used almost universally, that is the one to go for.


  19. @ Michael Glass who wrote: “What is clearly worse is where different systems of measurement are used side by side.”

    As things currently stand in the UK, the introduction of bridge height signs with metric units as well as imperial units (admittedly one above the other and not side by side) sadly still ranks as progress in our transition to metric. I often wonder why we can’t just get imperial off road signs altogether after forty years of metric teaching in schools.


  20. The mess of using two sets of measurement units (I refuse to call imperial a system) was highlighted for me just yesterday. I am renovating a house for a couple to move into, and the wife texted me asking if I could measure the windows so she could order new blinds. I duly measured the windows, and texted her the dimensions in metres, to two decimal places. She texted back, saying could she have the dimensions in either cms [sic] or in inches. When I pointed out that you just multiply the metre dimensions by 100 to get centimetres, her response was ‘Your [sic] a genius!’ This lady is 29 years old, and I would have thought completely metric-capable. Obviously not.

    I do believe, as has been said so many times in MV, that having to use two sets of measurement units makes you less than adequately comfortable in either.


  21. Whenever you try to keep two different sets of units going in parallel, there is going to be trouble.

    For example, this is from an article I just read today in the World Post section of the Huffington Post regarding climate change in Tibet:

    “Those heatwaves have been linked to thinning snow cover on the Tibetan plateau. The plateau’s temperature has increased by 1.3C (34.4F) – three times the global average.”

    If they would just stick with metric, we wouldn’t see such a laughable “conversion” to Fahrenheit.


  22. @Ezra Steinberg
    “The plateau’s temperature has increased by 1.3C (34.4F) –”
    That is the best one yet. As the saying goes “You just made my day!”
    It proves the point that the reporters, editors, proof readers or whoever, have no idea what they are talking about.
    Huffington post is one of the best metric users and the other units in the article are metric only. So it begs the question “why?”

    One of the presenters on BBC SE news actually used hectares last night, quickly followed by the on site reporter that due-fully gave the figure in acres.
    Another strange fact about conversions, is that metric to old stuff is approximated, but converting to metric requires a ridiculous level of precision (i.e. a calculator).


  23. 1.3C (absolute) is 34.34F (absolute), so even though they obviously missed the different zero points as opposed to delta C vs. delta F this is STILL wrong, doubly wrong.

    Having two numbers enables backtracking, though, by the educated. Had it just been a Celsius number wrong, orr just Fahrenheit, wrong, there’d be no way for a reader to catch it without comparing it to primary sourc information. I can look at the two numbers, see the discrepancy, do conversions, and figure out which is the correct number and what the error was. The person converting doesn’t even know how rounding works. So not someone with math or science as a strong suit.


  24. Ezra,

    This is the reporter’s or editor’s addition to what the scientist would have said and was quoted. Many reporters and editors don’t know the difference between an actual temperature and a temperature difference.


  25. @ACWM
    Can you read? Do you understand what you are reading?

    Just to make it absolutely clear to you, the article was about temperature RISE, not absolute (although absolute would need to be Kelvin and Rankine surely?).
    That is the whole point of the posts, mixing relative with absolute.
    Absolute tripe.


  26. ACWM misses the point that if one single standard system is used there is no need for ‘backtracking’ or ‘comparing it to primary source information’ or to ‘do conversions’ or ‘figure out which is the correct number’. The standard (metric) number was obviously the correct one to anyone who has been reading anything about this subject over the last few years. Where education comes into it, it is in standardising on a single standard system, not in professing to have some skill in juggling between two incompatible systems.


  27. An arrogant response, as I have pointed out that a temperature conversion, with different zero points, is obviously different than converting the differential between degrees of temperature change. It’s been a long time since thermodynamics, but I learned this trap of *RANKINE* and Kelvin being the only scales that will work for the equation, many years ago, as a Sophomore. I’d hope the writer graduated high school. . .


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: