The idea that the "day" should be counted at sunset as in the Hebrew calendar is considered by some as "preposterous" on two basic grounds: First, they don't "buy" the argument that in Genesis 1, where it refers to "thus it was evening and morning, 1 day" to imply the solar "day" should end and begin at sunset, that is, the idea that darkness is first followed by daylight is just not palatable to them; and second, some argue that the day ending at sunset is a is only a "Babylonian" influence, thus -pagan - and should hold no merit.
Let's discuss this a bit.
It is certainly true that the Bible does not actually "say" that the day ends and begins at sunset. (Neither does the Bible say you must "sight a crescent moon" for the beginning of the month, or that "you have to wait for the abib barley for the 1st month" - but these are not our present topic.) And granted, history does seem to confirm that the Hebrew calendar was probably influenced by Babylonian thinking at least in the period of Jewish exile in the 6th century BCE. Indeed, the Babylonian calendar was a luni-solar calendar, as is the Hebrew calendar, and the names of the months in today's Hebrew calendar, for example, can be linked to Babylonian names. But ancient ties to the Babylonian calendar is no reason to argue against the authority of Hebrew calendar by citing a "pagan origin" any more than we should deny the authority of our modern Gregorian calendar because of its ancient ties to the Roman empire! After all, some of the month names in our modern calendar are named after Roman emperors! For example, Quintilis was renamed as "Iulius" (July) in honor of Julius Caesar in 44 BC, and Sextilis was renamed "Augustus" (August) in 8 BC! Yet, no one is running around denying the authority of the Gregorian calendar!
So let's stop looking at the sketchy history and just look at the "visual" or "observable" facts that give us a calendar. After all, everyone worldwide sees sunrise and sunset, sees the moon throughout the month, and can see the motion of both the sun and the moon against the background of stars. So it seems reasonable that everyone on earth, if faced with developing a calendar from scratch, all have the same original observables and no "outside" influence is needed! The only problem comes in how those observations are interpreted. The Babylonians clearly thought their "gods" influenced the sun, moon and stars, therefore there's a lot of "deity" and pagan influence in their calendar! But if you separate the "deity" and "myth" influence from the sun, moon, and stars, we all have the same tools - the same observables with which to derive a calendar.
The 24-hour "day"
While the topic at hand is to discuss why or how it is that the Hebrew calendar day ends and begins at sunset, we find few arguing that our modern Gregorian calendar day which begins, oddly, at midnight should be questioned! Have you ever wondered what AM and PM mean in reference to clock time? Just before midnight the time is called 11:59:59 PM, then all the sudden it is 12:00:01 AM. Similarly, 11:59:59 AM just before noon all the sudden becomes 12:01:01 PM. What does this mean?
It comes from ancient Rome! When the sun rose each day making its way across the sky, every day it crossed a point in the sky exactly between the eastern horizon and the western horizon. The sun was, for a moment, exactly on the "middle line" between the east and west, that is, on the "meridian". The meridian is the imaginary line running through the sky from due north to straight south which naturally divides the east from the west. So when the sun was "on" the meridian, it was unquestionably mid-day. The sun on the meridian therefore represents mid-day. If the sun had not yet reached mid-day the sun was said to be ante-meridian meaning before the meridian. Similarly, when the sun had moved past mid-day it was said to be post-meridian meaning after the meridian. This is where "AM" and "PM" comes from! Therefore the time of day was referred to in ancient Rome as the position of the sun "before or after meridian". In those days "3 AM" meant the sun was "3 hours before mid-day", but now, in our day, "3 AM" means it is "3 hours after midnight" and "2 PM" means it is two hours after noon.
Without going into detail, perhaps now you understand why midnight became the time when our modern calendar increments the date. If mid-day is noon, from its Roman origins, then it follows that mid-night must be simultaneously the end of the old day, and the beginning of the new day. Thus the date "naturally" changes at midnight because mid-day was identified as the "middle" of the day!
But think about this a minute. Even if you decide that noon is the exact middle of the whole day, if you did not have a clock by which you could tell when it was "midnight", how would you have any idea when the time of midnight came? You wouldn't! All you can do, which is the same thing man has been able to do from the beginning of humanity, is watch the visible signs! And until clocks came along, there were no good* "signs" of midnight! Without clocks, the only visible sign of "what time is it" are these: sunrise, sunset, and mid-day. Truly, think about this. If you did not have any knowledge of a "clock" to show you the "time of day", then you'd have absolutely no idea in what part of the night "midnight" actually happened. If midnight marks the change of date but you don't know when midnight is, then you don't know when the date changed! Thus, if you were establishing "when" your day should end and begin, you would not "choose" midnight as the determinant!
(*Truth be had, one can use the stars to tell the time of midnight, but since the stars move across the sky, moving "ahead" at the same time each night roughly 4 minutes per day, you need to be a pretty astute observer to use the stars for telling time and thus discern midnight. So technically the stars do provide a sign of midnight, but not a good one for common use.)
So it is not "natural" to conclude the "date changes at midnight". Without a clock, you really have no idea when midnight happened, so by using the available visual signs, you'd never "choose" to have the day-count change at midnight. That leaves only sunrise, sunset, or possibly "mid-day" as the potential "observables" to decide when the day should change. What do we do next?
Let's whittle it down!
Let's next take mid-day as the potential "event" when the day count would change in our calendar and rule it out. It does not work for the following two reasons: First, the "moment" when the sun crosses the local meridian is not directly observable, that is, without some sort of instrument it is very hard to tell the moment when that happens. So we are left again with the need for a clock or a fixed gnomon to tell when "mid-day", or "noon", happens and we are trying to comprehend this in an era when clocks did not exist. Second, even if we "solve the problem" of knowing when the sun has reached the meridian, thus making "the moment" known, it would be very awkward, in any society, ancient or present, to change the "date" at that moment! Think about that for a minute! If we changed the date at what we know today as "noon", it would wreak havoc! So we can eliminate mid-day, or "noon" as a candidate for when our calendar date should change. A society simply would not do that.
That leaves only sunrise or sunset. We have ample evidence that several ancient societies have sided with one or the other. Some would call "sunrise" the new day, while some would call "sunset" the new day. The reason sunrise is chosen in some societies is beyond the scope of this article, but usually, a "deity" or myth enters into the decision. Ignoring such outside influence, what could be the difference between sunrise and sunset which might help us understand which might be the "right" choice, and to understand the Hebrew calendar without "dismissing" it as of "pagan origin"?
The simple answer is found by continuing our look at the visual signs.
We must now look at the nature and the meaning of the "month". Again, without going into much detail, the very origin of the word month is from the time it takes the moon to complete one cycle of its circuit around the earth. There is no question about this. By simply watching the moon, day after day, year after year, you will unavoidably conclude that the moon returns to the same measurable phase in almost exactly 29.5 days. This is why any calendar which uses the moon to define the "month" will alternate months between 29 and 30 days in length as the natural way to handle the month as "whole days". And please note that all I am talking about are the visual sign of the month, available to anyone worldwide, and it has nothing to do with "gods" or constellations, or "condition of green crops", or any "pagan" meaning. Simply look up, watch the moon, and your "month" must be 29 or 30 days. It can't be avoided.
Now, match the moon observations to the position of the sun compared to the background stars, and your "year" must be "12 months" because you will find that you see the moon renew itself 12 times each solar year. (The only problem you must deal with is that you would find the 12th month always ends "too early", that is, as the twelfth month ends there are still several days left before the year ends! You "balance" this problem by noting that you must "add" a 13th month once and a while to keep the sun, moon, and stars "in-sync". Explaining the need for and repetition of the 13th intercalary month is beyond the scope of this discussion). No "gods" or "pagan" festivals are needed to discover there are 12 months in the typical year, with a 13th month added every once and a while.
So now that we understand the month, let's zero-in on which is the proper visual sign to determine if the new day should be at sunrise or sunset.
If you have digested that all you need to do is watch the moon to determine the month, and you actually started watching the moon, you would quickly realize that the moon is always visible sometime during the day or night with one exception. That exception is that period of the month when the moon is very near the sun and is thus not visible, but is instead considered "in darkness". Of course "in darkness" is only a figure of speech. The only reason you can't still see the moon at that time is simply because it is too close to the sun and the sun is too bright! So the moon is actually "bathed in light", but you can see why it is considered "in darkness", after all, you can't see it! Nevertheless, that time when the moon is no longer visible, it turns out, is a quite natural discriminator of when the month can be said to have ended and the new month has begun.
You see, just as the Romans came to realize that the sun, at "mid-day" could be viewed as the exact 1/2 way point of the whole day and thus, "mid-night" defines the end and beginning of the day, the same observation is seen in the moon! Full Moon is the very visible, very obvious, very clear middle of the month, and that time when the moon is "in darkness" is the equivalent of "mid-night" for the sun, so very obviously, the time of the moon's "darkness", i.e., conjunction, MUST be when the month is both ending and being renewed! The new month ends and begins at conjunction! The moon itself says so!
It actually makes complete sense! Full moon is the middle of the moon's month, and conjunction is clearly the end and renewal of the moon's month. Many say that conjunction can't be used as the "sign" because it is not a "visible sign", insisting that a "sign" must be visible! In general, yes, a "sign" is something that you would expect to be visible, but in the case of the moon, its very absence is a powerful and obvious sign by the fact that it is NOT visible at that time! You may need to think hard about this! Remember, the moon is always visible sometime throughout each day, even if you must find it in the daylight, unless it is in the period of conjunction, thus the missing moon is the sign!
The argument against using conjunction as the time of the moon's renewal continues by saying "you can't know when conjunction happens". Honestly, the only people saying that are simply parroting what they've read or been told because all it takes to know when conjunction happened is to simply watch the moon and learn its signs! The moon tells you when it will be in conjunction. That phase called "conjunction", can be estimated pretty accurately by simply watching and recording when it is last seen and when it reappears and building knowledge month after month of what the moon shows you as the old month is ending. Also, you don't need to know the exact time of conjunction, rather, all you need to know is on what day conjunction happened. A skilled observer can look at the last crescent moon of the old month and tell you, without error, on which day conjunction will be! After conjunction, the moon itself is obviously "renewed", and now all you have to to is assign the correct "day" to your calendar.
But we are still working to decide which is the correct "event" to demark the "day", i.e., sunrise or sunset. There is one more key observation to make. Can you guess when the "renewed moon" first becomes visible?
The renewed moon is first visible at sunset! In fact, it is impossible to see the renewed moon at sunrise for a number of reasons. Therefore, it becomes a natural consequence of simply watching the sun, moon, and stars that you will make sunset the beginning of the new day since it is the ONLY time when an observation of the sun and the moon can be used to "mark" the same point in time - new day, new month! You can use the moon itself to confirm that the new month has begun, and you "confirm" that at sunset! (Don't get ahead of me - I am not talking about "sighting the new crescent".)
Though the moment of conjunction is in darkness and we agree you can't see it, you can know when it happened because, after all, you've been carefully watching the moon every day for centuries, and you can know which sunset after conjunction will be the "1st" sunset of the renewed moon. Since your current calendar month is ending with either the 29th or the 30th day (remember, you knew from the last, old crescent you saw if the current month needed a 30th day), that sunset after conjunction is the natural "right time" to watch for the sign of the renewed moon which will confirm the month.
So you watch at sunset at the end of the last day of the month, and watch for the new moon to reappear to confirm your calendar is correct. If you don't see the new crescent moon, at that last sunset of the current month, no problem, because you know - from centuries of observation - that sometimes the tiny crescent is "too new", too small, too faint to be visible as the sun sets on the last day of the old month and the new month begins. Nevertheless, if the crescent is not visible that sunset, the new crescent WILL be visible the next sunset as the 1st day of the new month ends. Thus, seeing the new crescent only confirms the new month! And, the size, shape and altitude of the new crescent at sunset when you do see it, reveals its "age", and you have validated your calendar!
If, on the other hand, you have decided to use sunrise as the demark of the calendar day, since it is impossible to see the new crescent moon as the sun rises, there is no "measure" for you to use to validate your calendar! How can the moon's reappearance at sunset (since you can't see the new moon at sunrise) mark the new month, when your "day" by a sunrise-day is already 1/2 over when you get your first chance to confirm if the moon has declared the new month? Using sunrise to mark the new day forces you to require some other mechanism to determine the month, but there is no other observable to provide any information about the month except the moon! Thus, you should now understand that in a lunar-solar calendar, sunrise is not a natural or wise choice to demark the day.
Our modern, civil calendar ignores the moon altogether, and the day is determined by neither sunset nor sunrise but by midnight as measured by a clock, and the months have a fixed number of days as determined long ago by the Roman emperors. None of our modern timekeeping alters the fact that when there were no clocks, and no Romans, the sun, moon and stars naturally divided the calendar into the day kept by the sun and months kept by the moon. And together, the moon and sun provided the natural choice of sunset as the demark of the day.
And notice how that does match Genesis 1.