There is no shortage of sites on the web which cover this ground to one extent or another, and it is for this reason that I've never dealt with Stardates in an article before. I'm doing so now because I've received several emails from people asking me if I have an explanation for the Stardate system and I was tired of typing it up over and over. It seemed a little too long for an FAQ answer, so I have made it an article.
During TOS Gene Roddenberry introduced several terms whose major purpose was to make it difficult to nitpick the show. For instance, the hand weapons carried by the crew were originally called lasers. Gene was concerned that since lasers are real-world items, viewers might object if his weapons were shown to do things that lasers could not do. Hence the introduction of "phasers" - fictional technology which could therefore behave however the writers wanted it to behave.
Gene was also concerned that the dialogue should not be too specific about when the show was set, in case the technology seen proved too ambitious or not ambitious enough for a stated year. To facilitate this, he introduced the concept of Stardates. The Stardate as created by Gene was little more than a semi-random string of numbers which allowed Kirk to state a specific date in his log entries, whilst keeping the timeline of the show undefined.
The problem comes with the "semi-random" nature. Take a look at the stardates of the episodes plotted against the episode number :
It is clear that there is a general correlation between the two - Stardates increase as episode number increases, and do so in an approximately linear way. So at first glance it seems that the Stardate system does indeed work reasonably well as a dating system.
However, there are problems. For instance episode 4 ("Mudd's Women") has a Stardate of 1329.1 whilst the earlier episode 3 ("The Corbomite Maneuver") has a Stardate of 1512.2. Episode 18 ("Shore Leave") has a date of 3025.3 whilst the next episode ("The Squire of Gothos") has a Stardate of 2124.5. Perhaps most importantly, episode 53 ("Patterns of Force") has a Stardate of only 2534 whilst those on either side of it have Stardates in the 4700s.
Still, there is no particular reason to presume that the episodes were numbered and aired in their "actual" order. So we can simply re-sort our list into Stardate order, renumber our episodes that way, and re-plot :
The lowest Stardate known is 1312.4 ("Where No Man Has Gone Before"), whilst the highest is 5943.7 ("Turnabout Intruder") - a span of 4631.3 units for the whole series. There is no real way to correlate this to a real world timescale because we don't know how far apart the episodes were set. We do know that the Enterprise was on a five year mission, however, so if we assumed that the episodes seen span almost the whole of this period, then one year would equal 926.26 units.
Further problems ensue when we introduce Stardates for events before the original series. In the episode "Where No Man Has Gone Before", Kirk views the service records of both Elizabeth Dehner and Gary Mitchell. Careful examination of Dehner's records on the screen shows that she was born on 1089.5 and is now 21 years old. Since we already know that this episode is set on Stardate 1312.4, we can see that 21 years equates to 222.9 Stardate units, so 1 year equals 10.61 units. This is a far cry from the near 1000 units to a year which seems to apply in TOS - at 10.61 units pre year Kirk's mission would have lasted for 87 years!
Mitchell's date of birth is 1087.7, and he is listed as 23 years old. This equates to 9.77 units per year, roughly equal to the value yielded by Dehner's file but also a far cry from the TOS figure.
I have seen some valiant efforts to come up with a workable way to interpret TOS dating systems. For instance, it has been suggested that the dates are unique to the logs of the ships making them and measure mission duration, with the first two digits representing the number of months which have passed since the ship embarked on its mission. This would work well for the Enterprise - the final episode of TOS ("Turnabout Intruder") has a Stardate of 5943.7 which would equate to just under five years into the ship's five year mission, perfect for the last episode. But it is hard to see how the ship's records can then list the date of birth of its crewmembers in Stardate format
One final problem comes in the episode "Mudd's Women". Kirk's opening log entry states that it is Stardate 1329.8 and that the ship is in pursuit of another vessel. Yet after the second vessel's occupants have been captured and brought on board, Kirk's next log entry indicates that it is now Stardate 1329.1. So time has apparently gone backwards!
In my humble opinion, any attempt to make TOS Stardates work in any logical way is doomed to failure.
Gene himself was once asked to explain the Stardate system. He replied :
|Roddenberry : ||"This time system adjusts for shifts in relative time which occur due to the vessel's speed and space warp capability. It has little relationship to Earth's time as we know it. One hour aboard the U.S.S.Enterprise at different times may equal as little as three Earth hours. The stardates specified in the log entry must be computed against the speed of the vessel, the space warp, and its position within our galaxy, in order to give a meaningful reading."|
And then promptly admitted that he had made this up on the spur of the moment, that he had no real idea what it meant, and that he would rather forget about the whole thing.
When TNG came along somebody - presumably Gene - decided to have a much more sensible, logical system of Stardates. To show that the new series was set later on than the original, all Stardates were five digit numbers rather than four. The first digit was a four, since the show was set in the 24th century. The next digit was a 1 since it was the first season of the show. The last three digits then rose from 000 at the beginning of the year to 999 at the end. So the first season of TNG spans 41000 - 41999, season 2 spans 42000 - 42999, and so on. Courtesy of the last episode in season 1 ("The Neutral Zone" ) we even have a real-world date for that season - in talking to some 21st century humans, Data states that "in your calendar it is two thousand three hundred sixty four."
So 41000 equates to the boundary between 31st December 2363 and 1st January 2364.
We know that 1000 units equates to one year, but we are uncertain exactly what is meant by a "year". A calendar year? A solar year? If so, whose calendar? Whose planet? What about leap years? None of these questions are addressed in any canon or official source that I know of.
We do know that TNG lasts for 7 seasons and that it spans 7 years, so apparently the years used are Earth years, or are from some calendar which is very close to it. For the purposes of this article I assume that 1000 stardate units equals exactly 365.25 days, and that this also equals exactly one Earth year. I further take one day to be exactly 24 hours of 60 minutes each. The latter assumptions are certainly untrue, but a fair approximation.
This means that 1 Stardate unit is equal to 0.36525 days, or 8 hours 45 minutes and 57.6 seconds. Alternatively, 1 hour would equal slightly over 0.114 Stardate units.
Whilst this system works very well for the most part, it has some interesting implications and some nits did creep in. For instance, since 2364 equates to 41000, then we can project backwards and discover that at the beginning of 1st January 2323, the Stardate was 00000.0. This is well after TOS, so the logical explanation seems to be that whatever system of Stardates was in play during TOS continued until 2322, and then a new system was introduced. Or 2323 might be thought of as some crucial dating point in Federation history, much as we divide dates into AD and BC. Neither assumption involves any problem with the TNG system, they just invite speculation as to what might have happened at that time.
Another issue which is not really a nit as such is that the idea of the first digit being a four because the show was set in the 24th century rather falls down after a while. Since a century has the years 0 - 99 in it (actually it has 1 - 100, but that's a different argument), at least two digits are needed to represent it. Yet the TNG system only allows one digit. Put succinctly, the end of TNG season 9 would have been 49999, but should season 10 then have gone to 410000 or 50000?
TNG didn't last to season 10 of course, but the stardate system used there was continued in Deep Space Nine and Voyager so essentially the dating system did continue that far and beyond. The writers chose the second option, and season 5 of Deep Space Nine featured a Stardate of 50049.3 ("The Ship"). Whilst this ensured that the system continued to work smoothly it does mean that the rationale of using 4 to represent the 24th century, whilst it still applies in the real world, should not be considered to apply within the fictional world of Star Trek.
Like TOS, the big nits with the TNG system tend to come when they give Stardates for events which happened well before the series. Since the TNG system only gives us 40 or so years prior to the series before we hit zero, and since many of the actors and characters shown are in their thirties or more even in season 1, past Stardates can easily push down towards that "zero barrier" and beyond. While there is nothing at all wrong with giving a Stardate of 01215.2, the writers really don't seem to like doing this for some reason. Nor have they ever settled on any Stardate for an event which would have happened before the 2323 zero barrier. The net result is that pre-TNG Stardates tend to become a little compressed.
I don't intend to try and list every example, but take for instance "Dark Page". Lwaxana Troi's marriage was stated to have taken place on Stardate 30620.1, which puts it in 2353. But this is clearly well after Deanna Troi was born, unless we assume she was eleven years old in season 1 of TNG. A Stardate in the 10000 range would be far more appropriate.
There are also other oddities. The Deep Space Nine episode "Second Sight" claims that the Stardate 47329.4 is the day after the fourth anniversary of the battle of Wolf 359, yet the Stardates for the episodes in which the battle occurred ("Best of Both Worlds" parts I and II) range from 43989.1 to 44001.4. The "Second Sight" Stardate would be some three and a half years after the battle, not four.
Another nit comes in the TNG episode "Data's Day". This describes a single day in the life of Data, and gives the Stardate as 44390.1. This would equate to about 22nd May 2367, yet Data says that the ship is preparing to celebrate the Hindu festival of lights, which comes in December.
None, really. No Stardate system makes total sense, but whilst the TOS system is so haphazard that it can hardly be called a system at all, the TNG system does at least have the advantage that you know what it is supposed to be doing. For giving us at least that much the creators deserve some kudos.