Britain is currently adopting European codes for structural design and allowing British Standard codes to lapse. This raises the issue of the preferred decimal marker – should imperial or continental practice be followed?
Mike Bather, who is a lecturer at the University of Bolton, writes in ‘The Structural Engineer’ of 19 July as follows:
“We as a nation have been driving on the left for far too long. The rest of Europe drives on the right and so should we. I propose we make the changeover gradually over the next the next 3 or 4 years.
Or rather, should I say that we, in the UK, make use of a decimal point in our calculations and when we come across a big number, we use commas to separate the thousands. The rest of Europe uses commas to separate whole numbers from decimals and does not bother to separate the thousands. The Eurocodes which are now our Eurocodes follow this continental approach. Thus we have the potential for an embarrassing mistake or worse.
To compound this, the Eurocodes contain many factors which are new to all of us. I have no idea whether lambda bar w should be 1.450 or 1,450. I am a little concerned that I may make an error one day confusing a point with a comma.
I appreciate that there is a world beyond Europe and, for instance, American engineers appear to follow our current conventions. In fact, the Americans are even more quaintly British than we are as they generally have stuck with pounds, feet and inches and no doubt rods, poles and perches as well. This just serves to complicate matters as we cannot align ourselves with Europe and America at the same time.
Speaking of Americans, they understand the cost of trivial mistakes. ‘People sometimes make errors,’ was how NASA explained the loss of their Mars Climate Orbiter spacecraft in 1999. This cost them over $300 million. The error in this case was a simple mix up over imperial and metric units; just the kind of slip that any engineer could make and any child could understand; just like confusing a comma with a decimal point.
Back in the UK, there are two camps developing. In camp one are the old engineers who love their decimal points. They would point out that most (but not all) Eurocode textbooks … stick with the decimal point. This camp includes the engineers who, if truth be told, would still rather be using BS449 and feet and inches.
(Ed. BS449 is a design code for structural steel, first published in the 1930’s)
In camp two are the new wave of students and graduates who are cutting their teeth on the Eurocodes. They are ganging up with the growing band of continental engineers practicing in the UK who think that commas are all you need. This camp includes people who can not write a proper number one and think that a seven needs a cross through it.
I foresee strife in our structural design offices and even worse, some engineers could possibly make some mistakes, confusing commas and points. As a lecturer in the UK (teaching home and international students), I ask for guidance: ‘Is it now time to stop using commas to separate thousands and to start using commas in place of decimal points or should I tell my students that they can drive on whichever side of the road they like?”
UKMA is considering publishing a Style Guide reference card in durable format, the purpose of which will be to help writers who would like to use metric measurement units correctly but are not quite sure of the rules. Where practices are internationally agreed, it is likely the Style Guide will recommend them, including the use of a space as a thousands marker. Where there is no international agreement, and this includes the preferred decimal marker, the editors are likely to give ease of use by the intended readers a high priority.
Metric Views 
You *never* ever use commas to separate thousands. The ISO standard is for spaces (i.e. thin-spaces) or no seperator at all. Use a comma or a point; I don’t care and it doesn’t matter as long as you don’t use commas to separate thousands.
LikeLike
A few comments on Mr. Bather’s article from an American perspective.
The SI Brochure (and NIST SP330) are quite clear that, depending on country, either the dot or comma may be used as a decimal marker, and therefore neither may be used as a thousands separator. Only the space (preferably thin space) may be used in this role. I would expect an engineer who is properly educated and working in SI units to be aware of this. NOTE: The US preference for decimal dot is documented in NIST SP330 which “interprets” the SI for the US; SP330 is available as a free pdf download.
Certainly the financial community does not use this standard for financial data and an engineer who works in Imperial or Customary may not regard this as applying to him. I believe all or nearly all English-speaking counties use the decimal dot, so the UK should also consider Canada, Australia, and others, not just the US. (Personally, I hope you maintain compatibility with other English-speaking nations.) One possible issue is that in some computer fonts, the decimal dot (also the period or full stop in text) is not very prominent and may be missed; perhaps we need a dot “on steroids.”
The US is a mix of companies and engineering disciplines working in metric and Customary. I personally worked in a metric industry, so I only have general knowledge of Customary practice. To my knowledge, no one uses yards, poles, perches, chains or stones. Surveyors use decimal feet (to hundredths). Machinists use decimal inches (to thousandths or ten-thousandths). Carpenters seem to like mixed base (feet and inches and common fractions). Most engineers (and their calculators) abhor mixed base.
I disagree with the summary he attributes to NASA. The Mars Climate Orbiter became the Mars Climate Auger because one supplier violated the terms and conditions of the purchase order which required metric units. Neither the supplier’s internal QC staff nor NASA caught the noncompliance. I think all parties were embarassed and preferred to sweep it under the table (or the Martian dust). I worked in a harsher industry; we would have held the supplier entirely accountable for the mission cost.
On the proposed Style Guide, it is way too detailed, but you may wish to look at NIST SP811 which gives guidance for metric usage beyond that in the SI Brochure for US practice, or ANSI SI10. I’m not suggesting you need to be the same, but it may be a useful resource. (I would be very interested if there are any recommended differences in practice.)
LikeLike
Well, well, well ….
Here is a welcome opportunity to drive home the point of the value of international standards. Not that the current government will listen, but perhaps key business interests might.
It might also be useful to suggest that documents following the new standard include in their legend an indication that commas are used as the “decimal point” just as drawings say that (for example) all measurements are in millimeters.
LikeLike
Is this a moot point? I don’t pretend to fully understand these Euro standards. However, a little Googling suggests that they are supposed to be preceeded by a national preface and adopted as a national standard. That national preface covers matters of national importance such as snow load requirements, but could certainly declare the decimal separator.
BS 449 is obsolete and long withdrawn, replaced apparently by BS 5950. However, per Wikipedia, in March 2010, BS 5950 was replaced by BS EN 1993. I assume that is the British national version of the Euro standard EN 1993. Assuming that understanding is correct, it would explain why Mr. Bather is seeing “Euro standards” with decimal points.
LikeLike
@John Steele
I was puzzled by your comment that our proposed Style guide is “way too detailed”. As you presumably haven’t seen our current draft (which is work in progress), I wasn’t sure what you were referring to. Can you clarify?
I also wasn’t sure how to take Mr Blather’s article, as he didn’t seem to be familiar with the BIPM brochure, and I had to check that it wasn’t 1 April.
LikeLike
@Erithacus,
Sorry, I was apparently unclear. I meant that SP811 is way too detailed, but worth looking at. In particular, they have a checklist for authors.
Obviously, it has things like American spelling (liter, meter) that I am not recommending you adopt, but parts of it are worth a look.
LikeLike
Both the National Building Code of Canada (published by the National Research Council of Canada), and the Handbook of Steel Construction (published by the Canadian Institute of Steel Construction), use the space as a thousands separator and the full stop/period as the decimal marker. In neither of those publications will you see a comma used in numerical values.
If I am reading anything from Europe (and possibly South Africa, Australia?), when I see a comma, I look first to see if it looks to be a decimal marker. Only in material from the US and the UK do I assume that the comma likely will be a thousands separator. Having said that, most newspapers and popular press in Canada still uses the comma as the thousands separator.
LikeLike
The article of Mike Bather, in ‘The Structural Engineer’ is a piece of ad hominem attacks! I agree with my country man John Steele. The standards of NIST and ISO are simple enough to work with when reading a document knowing the country of origin, for example South Africa or the United States when looking at expressed numbers e.g. 10,5 kg. Also seeing large numbers written as 1 000,00 is simple enough to read.
As a side issue, I see in retail many clothing items sold for the North American market, labels containing three languages French, English, and Spanish with sizes numbered with a , or . depending on the language.
Welcome to the Global Market Place!
LikeLike
Milliard is more appropriate to say 1,000,000,000.00 (or 1 000 000 000,00).
LikeLike
Computer languages are one of the issues that needs to be addressed regarding the use of the decimal separator. Most third generation languages such as C, C++, FORTRAN, Pasca, COBOL etc use the dot as the decimal separator, but GUI based systems such as EXCEL use an internal flag to identify the separator that should be used – in Windows this is set up on the “International preferences” page (where you can also set up the time format, currency format etc).
LikeLike
The solution to this compatibility problem between British practice and continental practice is to follow the BIPM guidelines on the use of the thousands separator and decimal marker. The BIPM says that thousands should be separated by a space and the dot or the comma should be used for the decimal marker. If the British switched from the comma to the space for the thousands separator, that would avoid any potentially embarrassing mistakes in pan-European collaborative projects that involve the UK.
LikeLike
I am Australian and Australia has officially always used the decimal dot rather than comma just like the British.
However, in the very first weeks of my professional life back in 1986 I made a very definite and conscious decision to use the comma, much to the annoyance of my colleagues. Here’s why. The dot is simply too easy to miss, especially in handwritten calculations. This is perhaps not so important today as in 1986, but one can still confuse a decimal dot with a speck of dust on the page. I know of no other symbol so heavily laden with meaning (shift its position and you’re out by at least an order of magnitude) so easily confused with noise if written as a tiny speck. In signal processing terms, as a dot, its signal to noise ratio is far too low. A few numerical mistakes in my first hand calculations in my first job convinced me that drastic action was needed to prevent more. I began using a vertical bar stretching above and below the line of numbers, but then I recalled that many countries use the comma, and indeed the Australian drawing standards at that time recommended a comma. Suddenly the (now changed) Australian drawing standard made sense to me, and I’ve used a comma ever since.
It’s true that one then has the problem of what symbol one uses for co-ordinate separators in vectors, sets, function argument separators and other mathematical notation. However, I’ve shifted in my career to a more theoretical physics / mathematics focus, and one simply doesn’t write numerical vectors, or sets or anything like this in practice. When one needs to write down components of a vector, or of a set, they are always pro numerals or other algebraic symbols. You don’t write down finite precision numbers in vectors in practice. I only very recently came upon a clash in Microsoft Excel, where Microsoft now sensibly gives you the choice as to which symbol to use for the decimal marker, and Apple Mac OS-X lets you make that choice for your whole computing world. So you need another symbol for a function argument separator in Excel formulas. For that, Microsoft uses a semicolon if you choose the comma as your decimal marker. That’s a little new and weird to me, but it works.
So in summary, I still believe the bigger, clearer comma to be superior to the dot, simply from a (human brain) signal to noise ratio standpoint.
LikeLike
I am Australian and Australia has officially always used the decimal dot rather than comma just like the British.
However, in the very first weeks of my professional life back in 1986 I made a very definite and conscious decision to use the comma, much to the annoyance of my colleagues. Here’s why. The dot is simply too easy to miss, especially in handwritten calculations. This is perhaps not so important today as in 1986, but one can still confuse a decimal dot with a speck of dust on the page. I know of no other symbol so heavily laden with meaning (shift its position and you’re out by at least an order of magnitude) so easily confused with noise if written as a tiny speck. In signal processing terms, as a dot, its signal to noise ratio is far too low. A few numerical mistakes in my first hand calculations in my first job convinced me that drastic action was needed to prevent more. I began using a vertical bar stretching above and below the line of numbers, but then I recalled that many countries use the comma, and indeed the Australian drawing standards at that time recommended a comma. Suddenly the (now changed) Australian drawing standard made sense to me, and I’ve used a comma ever since.
It’s true that one then has the problem of what symbol one uses for co-ordinate separators in vectors, sets, function argument separators and other mathematical notation. However, I’ve shifted in my career to a more theoretical physics / mathematics focus, and one simply doesn’t write numerical vectors, or sets or anything like this in practice. When one needs to write down components of a vector, or of a set, they are always pro numerals or other algebraic symbols. You don’t write down finite precision numbers in vectors in practice. I only very recently came upon a clash in Microsoft Excel, where Microsoft now sensibly gives you the choice as to which symbol to use for the decimal marker, and Apple Mac OS-X lets you make that choice for your whole computing world. So you need another symbol for a function argument separator in Excel formulas. For that, Microsoft uses a semicolon if you choose the comma as your decimal marker. That’s a little new and weird to me, but it works.
So in summary, I still believe the bigger, clearer comma to be superior to the dot, simply from a (human brain) signal to noise ratio standpoint.
LikeLike
One problem with using a decimal comma in the UK is that the comma is already widely used for separating groups of figures into thousands. Other countries which use a decimal comma leave a blank (hard) space to separate figures into thousands, thus avoiding the problem. Having said that, the EU has always used a decimal comma in figures, sums of money, etc. in its publications (in particular in the Official Journal of the EU) and the UK authorities have never had any problem in understanding that usage.
LikeLike