Standardising Time - An Essay

Foreword

This is a thought that has been rattling around inside my head for some years now, but I finally put fingertips to keyboard, so to speak, and here is the result. It was first inspired by my phyics teacher at college, Martin Crabtree, who mentioned the concept of a ten-day week, and I built upon the idea. Although I freely accept that this is probably not a new or original idea, the concept and words here are entirely from my own mind. However I did do a little research concerning leap years - the wepage I read is hyperlinked below in the relevant place. You are free to distribute this document freely (see Appendix A for information about saving and printing), but only as the complete document including header and footer, and also provide my URL (www.leevclarke.net) with it. Finally, if you're interested in further reading about time, you might want to try this as a starting point.

Standardised Time

Have you ever tried teaching a child how to tell the time? Or do remember what it was like trying to learn it yourself? It's easy enough on a digital clock or watch - if the display reads 3:45 then the time is three forty-five - but what about the traditional analogue design? This of course comprises a second hand, a minute hand and an hour hand. When the second hand completes a full revolution, that's a minute. Each revolution of the minute hand takes an hour. So each time the hour hand completes a full sweep of the clock face, that must be one day, right? Wrong! That's only half a day - the hour hand has to go around twice per day. If the hands point to twelve o'clock, is that midday or midnight? They are opposites of each other, and the fact is that we don't know which it is, just by looking at the clock face. This is somewhat of a problem, because the whole point of a clock is to tell you the time.

We all know what the face of an old analogue clock looks like. So what can be done to improve it? Well we could start by solving the issue I have just detailed, and make the numbers go from 1 to 24 instead of 1 to 12 twice - this would make the clock look something like the diagram A on the right, which shows 6:00am. To make this a workable clock, all we then need to do is to adjust the mechanism of the hour hand to make it travel around the clock at half the speed, so that one rotation takes a full day - this is easy to accomplish.

If you think about that, then at half an hour past midnight (12:30am under traditional time keeping) then under the new system, that would be called "half past twenty-four", for example. This in itself is silly - since we are half an hour into a new day, why are we still referring to it as half past twenty-four, i.e. yesterday's hours? To overcome this, we simply need to rename the hour of twenty-four as the hour of zero, so that time "resets" to zero at midnight, the start of a new day. Under this system, 12:30am would become "half past zero". As ridiculous as this might sound, I will justify the move in the next paragraph. This puts the face of the analogue clock as per diagram B below, and is the system I will refer to from here on as Standardised Time, or ST.

So what have we accomplished by this? Well firstly we have made the clock so that one rotation of the hour hand takes one day, so for example if the hour hand is one third of the way around the clock then we know that we are one third of the way through the day, i.e. 8:00am. This makes the hour hand work in a similar way to the minute and second hands. My diagrams both show six o'clock ST, or 6:00am under the old system, which is one quarter of the way through the day. Secondly, we have brought the analogue representation of time more in line with digital time on the twenty-four-hour clock - midnight is 0:00, or zero o'clock under Standardised Time. Midday is of course still twelve o'clock, since it is halfway through the day.

The position of the second hand tells us how far through the minute we are, the minute hand tells us how far through the hour we are, and the hour hand tells us how far through the day we are. In summary, the hands now tell us all the information we need to determine the exact time, accurate to one second, without the need for any additional information such as am or pm. One important consequence is that the twelve-hour-clock is rendered obsolete, but timekeeping is made much more comprehensive as a result.

Metric Time

While we're on the subject of standardising time, don't you think it's a bit strange that we have sixty seconds in a minute, and sixty seconds in an hour? These conventions come from the Babylonians who used sixty as the base of their numbering system - so the convention was fine for them. But since we in Britain now use the decimal system for money (pounds divided into one hundred pence), mass (the SI unit is the kilogram, divided into one thousand grams) and so on, why wouldn't we configure time to work in the same way? The answer if of course that no-one has tried to accept it as a new system - metric time is actually much better than imperial.

Firstly, how many seconds are there in a day? So that's 60 times 60 times 24. Now come on, don't go reaching for that calculator - just do it mentally. Finding it difficult? Point proven! The answer actually comes to 86,400 but I needed a calculator to do it too. Now suppose that we redefined time so that a day is divided into 10 hours, each hour is divided into 100 minutes, and each minute is divided into 100 seconds. Now how many seconds are there in a day? Easy - it's 100,000. That you can do in your head, no problem. Now suppose that under the new system I'm talking about 376 minutes - how many hours is that? 3.76 - another easy answer. So now you're getting the idea of why metric time is so much easier to work with. If the time is 5:87, how far through the day are you? 58.7% of course.

Now how do the lengths of these new units compare? Well since we have 86,400 imperial seconds equivalent to 100,000 metric seconds, it follows that a second will be slightly shorter, but not much. So a period of 10 metric seconds will take 8.64 imperial seconds, and you will not notice much difference. One minute on the other hand is significantly longer - it will be equivalent to 86.4 imperial seconds. Similarly one hour will take 8,640 imperial seconds, or 2 hours and 24 minutes in old units. With these latter two, it's simply a case of getting used to the new time-periods. This might seem like a mean feat but as I mentioned before, we in Britain have successfully achieved this with currency and mass, so why not time as well? We could even give the new time units completely different names to avoid confusion. If you're interested in finding out more about how metric time relates to imperial time, check out Appendix B for instructions on conversion of times.

There is a major issue that I should point out here, and that is that some well-known physical constants are defined according to the traditional definitions of time. The speed of light is one example - it's about 300,000,000 metres per second (to one significant figure). All these constants would need to be recalculated, but that is a simple matter of multiplying the old value by 0.864 (the number of imperial seconds in a metric second), or some other number to convert the units. While this might seem daunting, it is once again something that could be achieved. In my own diary I have tables that convert between imperial and metric units of distance, mass, volume, etc. Even these will not be needed in a few years because no-one will use imperial units any more. The same could be done for time units.

This is where we start dealing with time units larger than a day. A week for example, is traditionally 7 days including a 2-day weekend - which accounts for approximately 29% of the week. Now suppose that a week was defined as 10 days, including a 3-day weekend? That means that the weekend now works out to be 30%. In other words, the number of days lost as weekends in a year would be practically unchanged from the imperial system, and every weekend would be like a bank-holiday weekend! In fact the proposed metric system would give us about 5 more days off per year - perhaps these could be accounted for by abolishing bank holidays, and the whole year would balance out once again.

The only major obstacle comes at this point - defining a year. The two fundamental time lengths we cannot define are a day (the time taken for the planet to complete one rotation on its axis) and a year (the time taken for the planet to complete one full orbit around the Sun) - these are properties of our planet. The day was easy to divide down, but a year is not a whole number of days - it is approximately 365 and a quarter days. If we define a year as containing ten months, then that implies that each month would contain an average of 36.5 days each. This would leave us an extra quarter of a day per year to account for, which could be solved by adding an extra day on every four years - i.e. we would have a "leap year" every four years, as we do already. There are actually other provisions for which years are leap years, but in any case the current conventions about adding these extra days could still be used in the metric calendar.

So how do we make each month have an average of 36.5 days? Well that's easily solved by having alternating months with 36 and 37 days - five lots of these in a year makes a total of ten metric months, or 365 days as required! The extra day in a leap year would probably be tagged onto the end of a 36-day month. I would put it at the end of the year - in fact the extra day was originally at the end of the year under the Roman Calendar at the time (check out the link in the previous paragraph if you're interested in more details about how the leap year system was devised). In other words, the calendar would look something like this...

The days of weekends are highlighted with exclamation marks (for those of you reading in black and white). The very last day of the year (highlighted with an asterisk) would only occur in leap years - i.e. every four years there would be a 37th of J. Of course the ten months and ten days of the week would need to be renamed, although I have not take the liberty of doing this myself. It would be a good idea though to make sure that each day of the week begins with a different letter, so that they could be referred to by just one initial (there is a particular example of why that would be a sensible move detailed below, under "Dates - A Related Issue"). Also, the months of the year could be just numbered instead of named - that way we wouldn't have to remember the names of them. I understand that some people would dislike that idea, but it's a thought.

Another obstacle that needs to be addressed is the fact that many people would perhaps not like working for seven days in a row. This is of course an important point. One solution to the problem could be to divide the weekend into 2 days and 1, instead of three days. This could then be used to divide the working week into 4 days and 3. In other words, the week would become this: 4 days working, 1 day off, 3 days working, 2 days off. In this way, we would still get a 2-day weekend at the end of the week, as well as a day off in the middle, and we would never work for more than 4 days in a row! A two-part week sounds much more attractive to me anyway. And so I introduce my final draft of the metric calendar:

This may not look as tidy as the first calendar, but I think that it may well make for a more attractive schedule. Now consider this: if you were to take a two-week holiday under the current system, you miss ten working days, right? Now count the number of working days in the period I have highlighted with hash signs. That's right - there are exactly ten! In other words, the traditional fortnight's holiday could be replaced under the metric system by a pattern such as that highlighted, and you miss the exact same number of working days.

Global time-zones would presumably have to be reconfigured - would there be ten or a hundred? Or maybe some number in between? Another issue is the fact that the US still uses imperial units (in some places they are even referred to as American Units). America is a very large place, and it's hard trying to convince the general public that it's worth making the switch to metric units - it might well seem very daunting to them. On the other hand, maybe they could adopt metric time at the same time as all the other metric measurements such as mass, distance, etc. That way they perhaps wouldn't notice it so much, and it may well work to the advantage. Convincing the Americans may turn out to be the biggest problem of all. To begin with at least, we need to start by introducing metric time into Britain and Europe - after all, it has to start somewhere.

Dates - A Related Issue

While I'm waiting for the rest of the world to accept the metric time system, there is another issue I have something to say about - and that is the way in which we write dates. In Britain the convention used to be DD/MM/YY. With the famous "Millennium Bug", this has now been modified to a four-digit year, i.e. DD/MM/YYYY. However there is a different convention in the US, and that is MM/DD/YYYY, i.e. to write the month first, then the date, then the year. While this is actually even less logical than the British system, it turns out that in fact neither system is satisfactory. This first came to my attention when I wanted to sort files on my computer by date - and not by date last modified, so I couldn't use Windows' built-in "sort by date" feature. Something else was necessary.

I had to think about this a little more laterally. What I needed was a sort of code at the beginning of the filename, and then just sort the files by name. I needed the files to be sorted first by year, then by month, then by date - so that is the system I used. I actually ended up using a period (.) instead of the slash (/) to separate the numbers, so the system now looks like this: YYYY.MM.DD which of course works perfectly. Now this leads to the question of why this convention is not used the world over - it would solve the problem that currently exists, and that is that sometimes it's impossible to tell whether a date is written using British or American convention. For example, consider this: 5/3/1989. Is that the 5th of March (British) or the 3rd of May (American)? The new system of date-writing would clarify all those as well.

I am aware that at least one website already uses this method of recording dates. So why doesn't everyone else make the switch too? In fact it is not necessary to use a separator at all, although we as human beings would probably want one - a convention would need to be decided as to what character should be used. But actually any date can be written as an 8-digit string (YYYYMMDD) where a leading zero is included on dates and months that are single-digit. For example, the date I gave earlier would be written 19890305 (i.e. the 5th of March, which was the correct answer above). This would not present a problem even under the metric calendar, since no month or date is ever more than two digits. The final result is that an alphabetical sort is now equivalent to a chronological sort, with each file being intrinsically stamped with the required date. Any text (e.g. a more comprehensive filename) can come after the date, with no effect on the order of the files - unless two or more have the same date, in which case the relevant time could be tagged on immediately after the date. For example a file might be called "2001.08.27.14.37 RECORDS", meaning a file called RECORDS, with date 27th of August 2001 and time 2:37pm (I'm using the familiar imperial time to demonstrate the date system). This would give a true chronological sort - exactly as required.

The only time this system would be unsatisfactory would be when the year is more than four digits (i.e. from the beginning of the year 10000). This will not be for nearly 8,000 years, and while that in itself is not a justification for dismissing the issue, the fact is that there might be a new system for counting years by then. The proposed new-format date might also have a letter tagged onto the end of it to tell us the day of the week. This is purely for clarification of the date, to act as a sort of "checksum" to help make sure that the numerical date is correct, and would still keep the files in the correct order. For example, today's date as I write this would be 20010905W, meaning Wednesday the 5th of September, 2001. This is why I mentioned before that it would be a good idea for each day of the week to begin with a different letter - under the current system there would be confusion between Tuesday and Thursday, and Saturday and Sunday.

In summary, the old-fashioned conventions of writing dates (which are different in Britain and the US) may have been adequate in the past, but since we are now well into the digital age and the seams between nations are dissolving, a more satisfactory system is undoubtedly required. The best convention for writing dates is YYYY/MM/DD - although the separator character is debatable, and is not actually required at all. The initial letter of the day of the week could be tagged onto the end to clarify the date, but that would be best under metric time with each day having a different initial letter. Since this system works very well for computer dates, it would be most beneficial to us to write dates like that all the time - this would provide a seamless join between computers and "real life" conventions, eliminating confusion.

Appendices

Appendix A - Saving & Printing

The best way to save this document (like any webpage) with Internet Explorer is to click the File menu and select "Save As...". Save it as the filetype "Web Archive, single file (*.mht)". This saves the code as well as the graphics in one file, so you don't have to worry about transporting around an HTML document and a folder full of files - and you can also freely rename the single file. It takes up slightly more disk-space to do it like this, but it is well worth it in my opinion.

To print this document correctly, I recommend printing in landscape orientation to fit the calendars in properly. To do this, click the File menu and select "Page Setup...", then under "Orientation", select "Landscape" and click OK. Then click "Print" on the File menu to print the document.

Appendix B - Calculations

If you're interested in knowing how to convert from your standard imperial time to metric, then just follow these steps:

• Start with the time as digital format under the twenty-four hour clock, e.g. half past five pm would be 17:30.
• Multiply the hour by 60, then add the number of minutes. This tells you the total number of minutes into the day you are, i.e. since midnight (if you want to be more accurate, divide the number of seconds by 60 and add that on as well). In my example, the answer is 1,050.
• Divide this number by 1.44 - this converts imperial minutes to metric minutes. In my example, this comes to 729.167 (approximately). It's easy after here, because you just have to move the decimal point around - that's the advantage of the metric system of units.
• Divide this number by 100 to convert to metric hours - 7.29167 in my example. The first two digits after the decimal point tell you how many metric minutes it is past the hour, and anything after that tells you the number of metric seconds into the minute you are, In my example, this corresponds to 7:29 and 16.7 seconds on the metric clock. In other words, you are 72.9167% of the way through the day!

Conversion of time periods is just as easy to do - just start off with the time period in hours and minutes, and follow from step 2 above.

Well that's it, you've read all the way to the end of my essay. :-) If you know of any further information about these topics that you think I may be interested in, then please do not hesitate to email me to tell me about it. In particular, I'm interested in checking out any applications that will automatically display my PC's clock in Metric Time.