The shifted day
There has been some discussion on the newsgroup alt.usage.english recently of the advantages and disadvantages of the 12-hour and 24-hour clock systems. While the 24-hour clock is certainly useful in contexts where disambiguation between “am” and “pm” times is essential (e.g. transport timetables), it’s still seen by many as an artificial system more suited to technical contexts, and some people have difficulty interpreting 24-hour times. In my experience 12-hour times are almost universally used in everyday speech, and for many purposes “am” and “pm” are unnecessary, as the time of day can be interpreted from the context (e.g. a shop that advertised its hours as “8-6” would be readily understood).
On the other hand, the 12-hour clock has some serious drawbacks, the main one being that noon and midnight aren’t readily distinguished; the convention is that “12pm” is used for noon and “12am” for midnight, but they could easily be understood the other way around. In addition a time like “12.30am Tuesday” is easily misinterpreted as meaning half past twelve on Tuesday afternoon, when it actually means half past midnight on what’s technically Tuesday morning, but many would consider to be Monday night.
In fact, both systems have a problem at midnight. The 24-hour convention is that “00:00” refers to midnight of the day just starting, and “24:00” to midnight of the day just ending, but this appears not to be widely used or understood. On reflection I’ve come to the view that the main issue may not be which system is used, but where the day starts. Officially a new day begins at midnight, but in practice activities occurring in the early hours of the morning are considered to be part of the previous day’s schedule.
So here’s a novel solution that may get round the problem: start and finish the day later. I’ll take four o’clock in the morning as a starting-point, since in my experience the previous day’s activities have finished by then for the vast majority of people, and the new day’s schedule hasn’t begun yet. It’s also before dawn all year round at latitudes where the bulk of the population live. I’ll demonstrate how this can work with both the 24-hour and the 12-hour clock systems (ignoring seconds for clarity).
The shifted day with the 24-hour clock
Let’s say we start the new system on Monday 16th July 2012. Times up to one minute to midnight will be unaffected; in the ISO standard notation, this would be written “2012-07-16 23:59”. One minute later, the time changes to “2012-07-16 24:00” (i.e. on Monday), which is already accepted as an alternative notation for “2012-07-17 00:00”(technically Tuesday). One minute after that, the time would be written as “2012-07-16 24:01” (still Monday), which would be equivalent to ISO “2012-07-17 00:01”. This would continue right up until “2012-07-16 27:59” (Monday), equivalent to ISO “2012-07-17 03:59”. Note that there’s no need for a change in the ISO standard representation; all that’s being proposed is an alternative notation for the same thing. Since times past 24:00 don’t exist in the ISO standard, there’s no ambiguity. (The ISO standard would be one day ahead of the actual date though.)
One minute later, the time changes to “2012-07-17 04:00” (Tuesday), identical with the ISO standard. We have moved from 27:59 on Monday 16th July to 04:00 on Tuesday 17th July, and the new day has started. There is no possibility of confusion, since four o’clock is unambiguously regarded as the start of the new day; the time “28:00” wouldn’t be needed. We now continue in line with the ISO standard through until one minute to midnight, and then repeat the cycle.
The shifted day with the 12-hour clock
This is slightly more complicated as we need to add a third suffix alongside “am” and “pm” to designate the hours after midnight. However, this will help to solve the problems mentioned earlier. For illustration I’ll use “nt” (for “night-time”).
We start at 11.59pm on Monday 16th July. Moving on one minute, we reach 12.00nt on Monday 16th July (24:00 in the revised 24-hour clock). One minute later, it is 12.01nt on Monday (24:01), and so on until 3.59nt on Monday (27:59). A minute after that, it is 4.00am on Tuesday (04:00). We then go right through the new day as normal until 11.59pm, and then repeat the cycle.
Note that there is now no ambiguity over the representation for noon and midnight. Midnight is unambiguously “12.00nt”, and either “12.00am” or “12.00pm” can be used for noon. Furthermore, there’s no confusion over times like “12.30am”, because they don’t exist. Half past midnight is “12.30nt” and half past noon is “12.30pm”.
There’s an even more ingenious idea which would eliminate the possibility that “12.00pm” might be read as coming straight after “11.59pm”. If the day were divided into three eight-hour periods, then we could have “4.00am-12.00am”, “12.00pm-8.00pm”, and “8.00nt-4.00nt”, with the option of using either designation at the junctions. Thus noon would be either “12.00am” or “12.00pm”; eight in the evening would be either “8.00pm” or “8.00nt”; and four in the morning would be either “4.00nt” or “4.00am”. Of course this would have the drawback of bringing in an unfamiliar representation for the late evening hours.
It won’t happen, of course. But why not?