# On Geomantic Cycles

A while back on the Facebook community I manage for geomancy, the Geomantic Study-Group, someone had posted a proposed method to obtain four Mother figures for a geomantic reading based on the time and date of the query.  The poster based this proposal off of the Plum Blossom method of I Ching, where (as one of several possible formulas) you take the date and time and numerologically reduce the numbers to obtain trigrams; in a sense, such a method could theoretically be done with geomantic figures, and so the poster called this a type of “horary geomancy” (though I’m reluctant to use that term, because it’s also used by Gerard of Cremona to come up with a horary astrological chart by geomantic means, as well as by Schwei and Pestka to refer to geomancy charts that have horary charts overlaid on top).  He proposed three methods, but they all revolved around using the time of the query in astrological terms.

The proposed idea went like this:

1. Inspect the planetary ruler of the hour of the query.
2. Inspect the planetary ruler of the weekday of the query.
3. Inspect the planetary ruler of the Sun sign of the query.
4. Inspect the planetary ruler of the year of the query.
5. Transform the planets above, “taking into account rulerships by day or by night”, into geomantic figures, which are used as the First, Second, Third, and Fourth Mothers for the resulting chart for the query.

Seems straightforward enough!  I mean, I’m already familiar with the basics of horary astrology, I keep track of date and time cycles according to Greek letters, and I’ve flirted with using the Era Legis system of timekeeping as proposed by Thelema, and it’s even possible to extend the planetary hour system into planetary minutes and even seconds; having a geomantic system of time, useful for generating charts, seems more than fitting enough!  Besides, there’s already a system of geomantic hours based on the planetary hours which can probably be adapted without too much a problem.

I was excited for this idea; having a geomantic calendar of sorts would be a fantastic tool for both divination and ritual, if such a one could be reasonably constructed, and better still if it played well with already-existing systems such as the planetary week or planetary hours.  That said, I quickly had some questions about putting the proposed method from the group into practice:

1. What about the assignment of Caput Draconis and Cauda Draconis?  Do we just occasionally swap them in for Venus/Jupiter and Mars/Saturn, respectively, and if so, how?
2. Each planet has two figures associated with it; how do you determine which to pick?  “Taking into account rulerships by day or by night” isn’t always straightforward.
3. How do we determine the planetary ruler of a given year?
4. Is it possible instead to use the already existing cycles, such as the geomantic hours of Heydon, the rulerships of the lunar mansions, or the Cremona-based or Agrippa-based rulerships of the signs?

When I raised these questions (and a few others), I didn’t really get anything to clarify the method, so this particular conversation didn’t go anywhere.  This is unfortunate, because these pose some major problems to using a strictly planetary-based method of coming up with a geomantic cycle:

1. The issues in assigning the nodal figures to the planets is the biggest issue.  They simply don’t quite “fit”; even if you reduce the 16 figures into pairs, it’s hard to get eight sets mapped into seven planetary “bins”.  We see this quite clearly when we look at Heydon’s geomantic hours, where the nodal figures are sometimes given to the benefic or malefic planets (though I can’t determine a method), and on Saturdays, two of the hours of the Sun are replaced by the nodal figures (which is, itself, shocking and may just be a typo that can’t be verified either way).  Unless you expand a cycle of 24 hours or seven days into a multiple of 8 or 16, you’re not going to end up with an equal number of figures represented among the planets.
2. Given that each planet has two figures (ignoring the nodal figure issue from before), you can decide that one figure is going to be “diurnal” and the other “nocturnal”, or in planetary terms, “direct” or “retrograde”.  Different geomancers have different ways to figure out which of a planetary pair of figures are one or the other, so this might just be chalked up to individual interpretation.  Still, though, when would such a diurnal/nocturnal rulership actually matter?  Finding the figure for a planetary hour, using diurnal figures for diurnal hours and nocturnal figures for nocturnal hours?  Finding the figure for a weekday, using the diurnal figure if daytime and the nocturnal figure if nighttime, or alternating whole weeks in a fortnightly diurnal-nocturnal cycle?  Determining what figure to use if the Sun is in Leo or Cancer?
3. Multi-part problem for the issue of finding the “planetary ruler of a year”:
1. By inspecting the mathematics of the different kinds of planetary cycles that are established in the days of the week and the hours of the day, we can extend the system down into the minutes of the hours and the seconds of the minutes.  However, scaling up can’t be done along the same way; what allows for the planetary hours to work is that 24 does not evenly divide by 7, nor 60.  Because there’s always that remainder offset, you get a regularly repeating set of planets across a long system that, when aligned with certain synchronized starting points, allows for a planetary ruler of a given hour or day.  However, a week is exactly seven days; because there is no remainder offset, you can’t assign a planet ruling a week in the same way.  If you can’t even cyclically assign a planetary ruler to an entire week, then it’s not possible to do it for greater periods of time that are based on the week.
2. There is no method of cyclically assigning a planetary rulership to a year the way we do for days or hours.  The poster alluded to one, but I couldn’t think of one, and after asking around to some of my trusted friends, there is no such thing.  You might find the ruler of a given year of a person’s life, or find out what the almuten is at the start of a solar year at its spring equinox, but there’s no cyclical, easily extrapolated way to allocate such a thing based on an infinitely repeating cycle.
3. We could adopt a method similar to that in Chinese astrology: use the 12-year cycles based on the orbit of Jupiter, which returns to the same sign of the Zodiac every 11.8618 years (or roughly every 11 years, 10 months, 10 days).  In such a system, we’d base the planet ruling the year on the sign where Jupiter is found at the spring equinox.  This is both a weird import into a Western system that isn’t particularly Jupiter-centric, and is not quite exact enough for my liking, due to the eventual drift of Jupiter leading to a cycle that stalls every so often.
4. It’s trivial to establish a simple cycle that just rotates through all seven planets every seven years, but then the problem becomes, what’s your starting point for the cycle?  It’s possible to inspect the events of years and try to detect a cycle, or we can just arbitrarily assign one, or we can use mythological calendrics (a la Trithemius’ secondary intelligences starting their rulerships at the then-reckoned start of the world), but I’m personally uncomfortable with all these options.
4. Different existing cycles, different problems for each:
1. John Heydon’s geomantic hours from his Theomagia (which are the first instance I can find of such an application of the planetary hours) are a mess.  Even accounting for how he reckons the figures as “diurnal” or “nocturnal” and their planetary rulers, the pattern he has breaks at random points and I can’t chalk it up necessarily to being typos.  Additionally, there are 168 hours in a week, but this doesn’t evenly divide into 16, meaning that within a given week in Heydon’s (quite possibly flawed) system of geomantic hours, some figures will not be given as many hours as others.  If we went to a fortnight system of 14 days, then we’d end up with 336 hours which is evenly divisible by 16 (336 hours ÷ 16 figures = 21 hours/figure), but Heydon doesn’t give us such a system, nor have I seen one in use.
2. The system of lunar mansions from Hugo of Santalla’s work of geomancy ultimately formed the basis for the system of zodiacal rulerships used by Gerard of Cremona (which I’m most partial to).  However, of the 28 mansions, seven have no rulership, and five are duplicated (e.g. mansions 25, 26, and 27 are all ruled by Fortuna Minor).  Moreover, this system of attribution of figures to the mansions is apparently unrelated to the planetary rulership of the lunar mansions (which follow the weekday order, with the Sun ruling mansion 1).  It may be possible to fill in the gaps by closing ranks, such that the unruled mansion 7 is “absorbed” by Rubeus which already rule mansion 6.
3. There’s another system of lunar mansion rulership assigned to the figures, described by E. Savage-Smith and M. Smith in their description of an Arabian geomancy machine relating to directional correspondences, which uses the similarities between graphical point representation of the figures and certain asterisms of lunar mansions to give them their correspondence.  However, it is likewise incomplete, moreso than Hugo of Santalla’s assignments, and is likely meant as a way of cementing geomancy into Arabic astrological thought (though the two systems do share three figure-mansion correspondences, but this might just be coincidental overlap).
4. Hugo of Santalla’s system of lunar mansions and geomantic figures was eventually simplified into a set of zodiacal correspondences for the figures, such as used by Gerard of Cremona.  I like this system and have found it of good use, but Agrippa in his On Geomancy says that those who use such a system is vulgar and less trustworthy than a strictly planetary-based method, like what JMG uses in his Art and Practice of Geomancy.  Standardizing between geomancers on this would probably be the riskiest thing, as geomancers tend to diverge more on this detail than almost any other when it comes to the bigger correspondences of the figures.
5. Even if one were to use Agrippa’s planetary method of assigning figures to the signs of the Zodiac, you’d run into problems with the whole “diurnal” and “nocturnal” classification that different geomancers use for the figures, which is compounded with the issue of nodal figures.  For instance, according to Agrippa, Via and Populus are both given to Cancer; Carcer and Caput Draconis are given to Capricorn; and Puer, Rubeus, and Cauda Draconis are all given to Scorpio.  I suppose you might be able to say that, given a choice, a nodal figure is more diurnal than the planets (maybe?), but how would you decide what to use for Scorpio, if both figures of Mars as well as Cauda Draconis are all lumped together?

In all honesty, given my qualms with trying to find ways to overlay planetary cycles with geomantic ones, I’m…a little despairing of the notion at this point.  The systems we have to base geomantic cycles on are either irregular or incomplete, and in all cases unsatisfactory to my mind.

Now, don’t get me wrong.  I have heard that some geomancers have used the geomantic hours to good results, but I’ve also heard that some geomancers can get the methods of divination for numbers and letters to work; in other words, these are things that everyone has heard of working but nobody seems to have actually gotten to work.  And, I suppose if you don’t think about it for too long and just take it for granted, perhaps you can get the geomantic hours to work!  After all, I’ve found good results with Hugo of Santalla’s figure-mansions correspondences, even if they’re incomplete and unbalanced, without anything backing them up.  (I never denied that over-thinking can be a problem, much less a problem that I specifically have.)

Further, I’m not saying that geomantic cycles don’t exist; they very likely do, if the elements and the planets and the signs all have their cycles in their proper times.  The problem is that so much of these other cycles we see are based on fancier numbers that are either too small or infrequent (4 elements, 7 planets) or don’t evenly divide into 8 or 16 (like 12 signs, 27 letters in an alphabet), or they simply don’t match up right.  For instance, it would be possible to create a new set of geomantic hours where each figure is present in turn over a course of 16 hours, then repeat the cycle; this leads to returning to the same figure at the same hour of the day every 48 hours, starting a new cycle every third day.  This doesn’t match up well with a seven-day week, but rather a cycle of two weeks (as hypothesized above, since 14 days = 336 hours, and 336 is divisible evenly by 16).  However, such a system would break the correspondence between planets and figures because of the “drift” between cycles of 16 and 7.

So…in that line of thinking, why not rethink the notion of geomantic cycles apart from tying them to planetary ones, and start from scratch?

We’re accustomed to thinking of magical cycles in terms of seven planets, but we could just as easily construct cyclical time systems in terms of four (which can be divided four ways within it), eight (divided into two), or sixteen units.

• Consider the synodic period of the Moon, which can be said to have eight phases: new, crescent, first quarter, gibbous, full, disseminating, third quarter, and balsamic.  We could attribute each phase two figures, and then sync the cycle to, say, the new moon (when the Sun and Moon are in conjunction) or to the first quarter moon (when the Sun sets as the Moon is directly overhead), giving a synodic month 16 geomantic “stations” each lasting about 1.85 days.
• Those with a neopagan background are used to thinking of the year as an eight-spoked Wheel, where the year is divided by eight sabbats, which are four quarter days (equinoxes and solstices) and four cross-quarter days; each period between one sabbat and the next could be split into a geomantic “season” lasting roughly 22 or (sometimes) 23 days long.
• Alternatively, a year of 365 days can be broken up into 22 “months” of 16 days each, leading to 352 days, meaning three or four intercalary/epagomenal days at the end of the year or spread around for, say, the quarter days.
• Within a single day from sunrise to sunrise, we can divide the day into four segments (morning, afternoon, evening, and night) divided by the stations of the sun (sunrise, noon, sunset, midnight), and each segment can be further subdivided into four geomantic “hours”, leading to a total of 16 geomantic “hours” within a day which would, assuming a day of equal daytime and nighttime, have each “hour” equal to 90 minutes.
• Years can be broken down into cycles of four years, every fourth year requiring a leap day; this could lend itself to a cycle of 16 years (one geomantic figure per year), or even to a cycle of 64 years (comprising 16 leap days), each of which can be used as a way to define larger-time cycles.

Such a four- or eight-fold division of time and space isn’t unheard of; we commonly reckon a year (at least in most Western Anglophone countries) as having four seasons, the Greeks broke up cycles of years into four-year Olympiads, the ancient Romans divided up the night into four watches (while using twelve hours for the daytime), and there are discussions of a Hellenistic system of astrological houses called the octotopos/octotropos system which uses eight houses instead of the usual 12, so it’s possible to dig that up and rework it to accustom a geomantic method where the number 16 could be applied to work better than mashing it onto a system where the number 7 is more prominent.  That said, finding such a system that’s thoroughly based on 4, 8, or 16 is difficult, as it’d be pretty artificial without including the moon (which repeats in patterns of 12 or 13) or whole number divisors of 360, and considering how thoroughly cultural transmission/conquering has established the 12-month year across most of the world, often obliterating and subsuming earlier systems that may not have left much of a trace.  But, again, if we’re gonna just up and make one from scratch, I suppose it doesn’t need to be grounded in extant systems, now, does it?  Even if it’s artificial, if it’s a cycle that works, such as by associating the different motions of the sun and sensations of the day with the figures, or by linking the changes in the seasons with the figures, then that’s probably the more important thing.

Unlike my older grammatomantic calendars, where the order of the letters provided a useful guide to how the system should “flow”, the geomantic figures have no such inherent order, but can be ordered any number of ways (binary numeral equivalence, element and subelement, planetary, zodiacal order by Gerard of Cremona or by Agrippa, within one of the 256 geomantic emblems, the traditional ordering of odu Ifá which we shouldn’t ever actually use because this isn’t Ifá, etc.).  Or, alternatively, new orders can be made thematically, such as a “solar order” that starts with Fortuna Maior at sunrise, continues through the figures including Fortuna Minor at sunset, and so forth.  This would be a matter of experimentation, exploration, and meditation to see what figure matches up best with what part of a cycle, if an already existing order isn’t used as a base.

I do feel a little bad at not offering a better alternative to the problem that the original poster on Facebook posed, instead just shooting it down with all my own hangups.  Over time, I’d eventually like to start building up a geomantic calendar of sorts so as to try timing things for geomantic spirits and rituals, but that’ll have to wait for another time.  Instead, going back to the original problem statement, how can we use time to come up with four Mothers?  Well, perhaps we can try this:

1. Consider four lists of geomantic figures: binary (B), elemental (E), planetary (P), and zodiac (Z).  Pick a list you prefer; for this method, I recommend the simple binary list (Populus, Tristitia, Albus…Via).  Enumerate the figures within this list from 0 to 15.
2. Look at the current time and date of the query being asked.
3. Take the second (1 through 59, and if the second is 0, use 60), minute (ditto), and hour (1 through 23, and if 0, use 24).  Add together, divide by 16, and take the remainder.  This is key 1.
4. Take the day of the year (1 through 365 or 366), divide by 16, and take the remainder.  This is key 2.
5. Take the year, divide by 16, then take the remainder.  This is key 3.
6. Add up all the digits of the current second, minute, hour, day, and year.  Divide this number by 16, then take the remainder.  This is key 4.
7. For each key, obtain the corresponding Mother by finding the figure associated with the key in the list you choose.

So, for instance, say I ask a query on September 25, 2017 at 9:34:49 in the evening.  According to the method above, starting with the actual math on step #3:

1. Since 9 p.m. is hour 21 of the day, 49 + 34 + 21 = 104.  The remainder of this after dividing by 16 is 8, so K1= 8.
2. September 25 is day 268 of year 2017.  The remainder of 268 ÷ 16 is 12, so K2 = 12.
3. The remainder of 2017 ÷ 16 is 1, so K3 = 1.
4. 49 + 34 + 21 + 268 + 2017 = 2389, and the remainder of this after dividing by 16 is 5, so K4 = 5.
5. Using the binary list, (K1, K2, K3, K4) = (8, 12, 1, 5), which yields the Mother figures Laetitia, Fortuna Minor, Tristitia, and Acquisitio.

While this is not a perfect method, since the number of days in a year is not perfectly divisible by 16, the possibilities of each figure appearing as a Mother are not exactly equal to 1/16, but the process is decent enough for pretty solid divination based on time alone.  Instead of using purely date/time-based methods, you could also use the birth information of the querent alongside the date and time of the query, use the figures for the current geomantic hour/lunar mansion/Sun sign of the Zodiac, or numerologically distill the query by counting the number of letters or words used or by using gematria/isopsephy to distill and divide the sum of the content of the query.  So, I a method like what the original poster was proposing could certainly work on strictly numerical principles alone, just not on the astrological or planetary cyclical methods proposed.

As for geomantic cycles, dear reader, what do you think?  If you were to link the geomantic figures to, say, the phases of the moon, the eight “spokes” of the neopagan Wheel of the Year, or the flow of light and darkness across a day reckoned sunrise-to-sunrise, how would you go about creating such a cycle?  Have you used the geomantic hours, and if so, have you run into the same problems I have, or have you used them with good effect, in lieu of or in addition to the normal planetary hours?

# Details on the Grammatēmerologion

Yes, it’s official.  I’m settling on the term γραμματημερολογιον grammatēmerologion as the official term for the lunisolar grammatomantic calendar, including its chronological ritual use to schedule magical rites and festivals.  Long story short, this is a lunisolar calendar that maintains the lunar synodic months of 29 or 30 days in a particular cycle of either 12 or 13 months for every year to keep track with the seasons and the solar year.  What makes this different is that the days of the lunar month, as well as the months and the years themselves, are attributed to the letters of the Greek alphabet, hence grammatomantic for their ritual and occult significations.  If for some reason, dear reader, you don’t know what I’m talking about yet, go read through those two posts I just linked and learn more.

At its core, the major use of the Grammatēmerologion system is to keep track of monthly ritual days.  Of the 29 or 30 days in a lunar month, 24 are attributed to the 24 letters of the Greek alphabet; three are attributed to the obsolete letters of the Greek alphabet that were phased out (Digamma, Qoppa, and Sampi); and the other two or three are simply unlettered days.  Each of the 24 letters of the Greek alphabet is associated with a particular elemental, planetary, or zodiacal force according to the rules of stoicheia, and by those associations to one or more of the old gods, daimones, and spirits of the ancient Greeks.  Thus, consider the second day of the lunar month; this day is given the letter Beta.  Beta is associated with the zodiacal sign Aries, and by it to the goddess Athena and her handmaiden Nike.  Thus, scheduling sacrifices and worship to Athena and her attendant spirits on this day is appropriate.  The rest goes for the other days that are associated to the 24 letters of the Greek alphabet.  The three days given to the obsolete letters are given to the ancestral spirits of one’s family and kin (Digamma), one’s traditions and professions (Qoppa), and to culture heroes and the forgotten dead (Sampi).  The unlettered days have no ritual prescribed or suggested for them, and the best thing one can do is to clean up one’s house and shrines, carry out one’s chores, and generally rest.

Given a calendar or a heads-up of what day is what, that’s all most people will ever need to know about the Grammatēmerologion system.  Anything more is for the mathematicians and calendarists to figure out, although there are a few things that the others should be aware of.  For instance, there’s the problem of figuring out what months have 30 days (full months) and what months have 29 days (hollow months).  Add to it, in order to maintain a link between the lunar months and the solar year, we need to figure out which years need 13 months (full years) instead of the usual 12 (hollow years).  There’s a method to the madness here, and that method is called the Metonic cycle.  The cycle in question was developed by the Athenian astronomer Meton in the 5th century BCE, and he calculated that 19 solar years is nearly equal to within a few hours to 235 synodic months of the Moon.  Meton prescribes that for every 19 solar years, 12 of them should contain 12 synodic months and seven should contain 13; there should be a full year of 13 months after every two or three hollow years of 12 months.  Likewise, to keep the lunar month fixed to the actual phases of the Moon, a hollow month of 29 days should follow either one or two full months of 30 days.

Now, I won’t go into all the specifics here about exactly what month in what year of the Metonic cycle has 29 or 30 days or the gradual error that accumulates due to the Metonic cycle; that’ll be reserved for another text and another time.  Suffice it to say that Meton was very thorough in developing his system of 19 years and 235 days, figuring out when and where we should add or remove a day or a month here or there, and I’ve used his system in developing a program that calculates what the lunar date is of any given Gregorian calendar date.  (If you’re interested, email me and I’ll send you the Python code for private use only.)  If you want to read more about the specifics of the Metonic cycle developed and employed in ancient Greece, along with other calendrical schemes that the Metonic cycle was based on and influenced later on, I invite you to browse the six-volume work Origines Kalendariæ Hellenicæ by Edward Greswell from the 1860s (volumes one, two, three, fourfive, and six).  Yes, this is a nasty endeavor, but hey, I did it, so you can too.

So, let’s take for granted that we have the Metonic cycle of hollow and full months and hollow and full years.  We have a cycle of 19 years that repeats; cool!  The problem is, where do we start the cycle?  Without having a start-point for our Metonic cycle, we don’t have a way of figuring out which year is which in the Metonic cycle.  In the post where I introduced the lunisolar grammatomantic calendar, I sidestepped this by using the same cycles as another lunisolar calendar that makes use of a system similar to (but isn’t exactly) the Metonic cycle, that of the Hebrew calendar.  However, after researching the differences between the two, I decided to go full-Meton, but that requires a start date.  This start date, formally called an epoch, would be the inaugural date from which we can count these 19-year cycles.  The question then becomes, what should that start date be?

Well, the structure of the lunisolar grammatomantic calendar is based on that of the Athenian calendar, which starts its years on the Noumenia (the first day after the New Moon) that immediately follows the summer solstice.  Looking back at history, I decided to go with June 29, 576 BCE.  No, the choice of this date wasn’t random, and it was chosen for three reasons:

• The New Moon, the day just before the Noumenia, occurred directly on the summer solstice.
• The summer solstice coincided with a total solar eclipse over Greece.
• This was the first year after the legislative reform of Solon of Athens in 594 BCE where the Noumenia coincided with the summer solstice so closely.

Thus, our first cycle of the Grammatēmerologion system begins on June 29, 576 BCE.  That date is considered the inaugural date of this calendrical system, and although we can track what the letters of the days, months, and years were before that, I’ve chosen that date to count all further dates from in the future.

Still, there’s a bit of a caveat here.  Recall that, in a 19-year cycle, there are 12 years with 12 months and seven years with 13.  12 is a nice number, but for the purposes of working with the Greek alphabet, we like the number 24 better.  Thus, instead of using a single Metonic cycle of 19 years, a grammatemerological cycle is defined as two Metonic cycles, i.e. 38 years.  Thus, in 38 years, there will be 24 hollow years and 14 full years.  At last, we can start assigning the Greek letters to periods longer than a day!  The 24 hollow years are the ones that have Greek letters, and these are given in order that they’re encountered in the grammatemerological cycle; the 14 full years, being anomalous, are left unlettered.

The only thing left now is to assign the letters to the months themselves.  In a year, we have either 12 or 13 synodic months, and that 13th month only occurs 14 times in a period of 38 years; we’ll make those our unlettered months.  Now, again, within a year, we only have 12 months, and we have 24 Greek letters to assign.  The method I choose to use here is to assign the 24 letters of the Greek alphabet to the 24 months in two successive years.  That means that, in the cycle of 38 years, the odd-numbered years will have month letters Α through Μ, and the even-numbered years will have month letters Ν through Ω.  This doesn’t mean that we’re redefining a year to be 24 (or 25) months, but that our cycle of associating the letters of the Greek alphabet makes use of two years instead of just one.  This is only cleanly possible with a dual Metonic cycle of 38 years, since a single Metonic cycle of 19 years would have both that final 19th year and the next initial first year both have month letters Α through Μ.

If you’re confused about the resulting system, I got your back.  Below are two charts I had already typed up (but really don’t wanna transcribe into HTML tables, although it feels awkward to take screenshots of LaTeX tables) that describe the complete system.  The first table shows what months are full and hollow within a single Metonic cycle of 19 years.  The second table shows what years and months within a dual Metonic cycle of 38 years get what letters.

Like I mentioned before, this is getting really in-depth into the mechanical details of a system that virtually nobody will care about, even if they find the actual monthly calendar useful in their own work.  Then again, I’m one of those people who get entranced by details and mathematical rigor, so of course I went through and puzzled this all together.  Ritually speaking, since we ascribe particular days to particular forces or divinities, we can now do the same for whole months and years, though with perhaps less significance or circumstance.

However, these details also yield an interesting side-effect to the Grammatēmerologion system that can be ritually and magically exploited: that of Μεγαλημεραι (Megalēmerai, “Great Days”) and Μεγιστημεραι (Megistēmerai, “Greatest Days”).  Because the day, month, and year of a given grammatemerological date each have a given letter, it’s possible for those letters to coincide so that the same letter appears more than once in the date.  So, for instance, on our epoch date of June 29, 576 BCE, this was the first day of the first month of the first year in a grammatemerological cycle; the letter of the day, month, and year are all Α.  In the second day of the second month of the first year, the letters of the day and month are both Β and the letter of the year is Α.  These are examples of a megistēmera and a megalēmera, respectively.

• A megalēmera or “Great Day” occurs when the letters of the day and the month are the same with a differing letter of the year.  A megalēmera occurs in every month that itself has a letter, so not in those 13th intercalary months in full years.  Because it takes two years to cycle through all 24 month letters, a particular megalēmera occurs once per letter every two years.
• A megistēmera or “Greatest Day” occurs when the letters of the day, month, and year are all the same.  A megistēmera can only occur in years and months that themselves have letters, so megistēmerai cannot occur in full years.  A particular megistēmera occurs once per letter every 38 years, but not all letters have megistēmerai.  Only the ten letters Α, Ε, Ζ, Κ, Λ, Ν, Ρ, Σ, Χ, and Ψ can receive megistēmerai due to the correspondence between the letters of the year and the letters of the month based on whether the year is odd or even.

In a sense, these are like those memes that celebrate such odd Gregorian calendrical notations such as 01/01/01 (January 1, 1901 or 2001) or 11/11/11 (November 11, 1911 or 2011).  However, we can use these particular dates as “superdays” on which any particular action, ritual, offering, or festival will have extra power, especially on the comparatively rare megistēmerai.  These days are powerful, with the force and god behind the letter of the day itself extra-potent and extra-important, and should be celebrated accordingly.  It’s similar to how the system of planetary days and hours work: yes, a planetary hour is powerful, and a planetary day is also powerful, and if you sync them up so that you time something to a day and hour ruled by the same planet, you get even more power out of that window of time than you would otherwise.  However, megalēmerai are comparatively common, with 12 happening every year, compared to megistēmerai, which might happen once every few years.

Consider the next megistēmera that we have, which falls on October 17, 2015.  In 2015, we find that June 17 marks the start of the new grammatemerological year; yes, I know that this falls before the summer solstice on June 21, but that’s what happens with lunar months that fall short of a clean twelfth of the year, and hence the need for intercalary months every so often.  The year that starts in 2015 is year 7 of the 69th cycle since the epoch date of June 29, 576 BCE.  According to our charts above, the seventh year of the grammatemerological cycle is given the letter Ε.  Since this is an odd-numbered year in the cycle, we know that our months will have letters Α through Μ, which includes Ε.  The letter Ε is given to the fifth month of the year, which begins on October 13.  We also know that the letter Ε is given to the fifth day of the month.  Thus, on October 17, 2015, the letter of the day will be Ε, the letter of the month will be Ε, and the letter of the year will be Ε.  Since all three letters are the same, this qualifies this day as a Megistēmera of Epsilon.  This letter, as we know from stoicheia, is associated with the planetary force of Mercury, making this an exceptionally awesome and potent day to perform works, acts, and rituals under Mercury according to the Grammatēmerologion system.  The following Megistēmera will be that of Zeta on November 25, 2017, making it an exceptionally powerful day for Hermes as a great generational day of celebration, sacrifice, and honor.

As noted before, only the ten letters Α, Ε, Ζ, Κ, Λ, Ν, Ρ, Σ, Χ, and Ψ can receive megistēmerai.  To see why Β cannot receive a megistēmerai, note that Β is assigned to the second year in the 38-year grammatemerological cycle.  Even-numbered years have months lettered Ν through Ω, and the letter Β is not among them.  This is a consequence of having the months be given letters in a 24-month cycle that spreads across two years.  We could sidestep this by having each month be given two letters, such as the first month having letters Α and Ν, the second month Β and Ξ, and so forth, but that complicates the system and makes it less clean.  Every letter receives two megalēmerai per grammatemerological cycle, but only these specified ten letters can receive megistēmerai; whether this has any occult significance, especially considering their number and what they mean by stoicheia, is something I’ve yet to fully explore.

So there you have it: a fuller explanation of the lunisolar grammatomantic calendar, known as the Grammatēmerologion system, to a depth you probably had no desire to investigate but by which you are now enriched all the same.  It’s always the simple concepts that create the most complicated models, innit?

# Lunisolar Grammatomantic Calendar

In my first post on grammatomantic calendars and day cycles, I hypothesized that it would be possible to a kinds of calendar suitable for assigning a Greek letter (and, by extension, the rest of its oracular and divinatory meaning) to a whole day without an explicit divination being done, similar to the Mayan tzolk’in calendar cycle.  I did this creating a solar calendar of 15 months of 24 days each, each day assigned to a different letter of the Greek alphabet in a cycle, and also extended it to months, years, and longer spans of time; its use could be for mere cyclical divination or for more complex astrological notes.  At its heart, however, it is essentially a repeating cycle of 24 days, plus a few correctional days every so often to keep the calendar year in line with the solar year.  Because of this, it is essentially a solar calendar, keeping time with the seasons according to the passage of the sun.

Awesome as all this was, it’s also completely innovative as far as I know; the Greeks didn’t note time like this in any recorded text we have, and it takes no small amount of inspiration from the Mesoamerican Long Count calendar system.  Wanting a more traditional flavor of noting time, I also hypothesized that it might be interesting to apply a grammatomantic cycle of days to an already-known calendar system used in ancient Greece, the Attic festival calendar.  In this case, the calendar system already exists with its own set of months and days; it’s just a matter of applying the letters to the days in this case.  No epoch nor long count notation is necessary for this, since it’s dependent on a lunar month a certain number of months away from the summer solstice (the starting point for the Attic festival calendar).  The primary issues with this, however, is that the Attic festival calendar is lunisolar following the synodic period of the Moon, so it has months roughly of 29 or 30 days, depending on the Moon.  This is more than 24, the number of letters used in Greek letter divination, and 27, the number of Greek letters including the obsolete digamma, qoppa, and sampi.  With there being only 12(ish) months in this calendar system, this is going to have some interesting features.  To pair this calendar with the Solar Grammatomantic Calendar (SGC), let’s call this the Lunisolar Grammatomantic Calendar (LGC).

So, to review the Attic festival calendar, this is a lunisolar calendar, a calendar that more-or-less follows the passage of the Sun through the seasons using the Moon as a helpful marker along the way to determine the months.  Many variations of lunisolar calendars have been created across cultures and eras, since the changing form of the Moon has always been helpful to determine the passage of time.  With the Greeks, and the Attics (think Athenians, about whom we know the most), they used the fairly commonplace system of 12 months as determined by the first sighting of the new Moon.  As mentioned, the start date for the Attic festival calendar was officially the first new Moon sighted after the summer solstice, so the year could start as early as late June or as late as late July depending on the lunar cycle in effect, making mapping to the Gregorian calendar difficult.  The names of the 12 months along with their general times and sacredness to the gods are:

1. Hekatombaion (Ἑκατομϐαιών), first month of summer, sacred to Apollo
2. Metageitnion (Μεταγειτνιών), second month of summer, sacred to Apollo
3. Boedromion (Βοηδρομιών), third month of summer, sacred to Apollo
4. Pyanepsion (Πυανεψιών), first month of autumn, sacred to Apollo
5. Maimakterion (Μαιμακτηριών), second month of autumn, sacred to Zeus
6. Poseideon (Ποσειδεών), third month of autumn, sacred to Poseidon
7. Gamelion (Γαμηλιών), first month of winter, sacred to Zeus and Hera
8. Anthesterion (Ἀνθεστηριών), second month of winter, sacred to Dionysus
9. Elaphebolion (Ἑλαφηϐολιών), third month of winter, sacred to Artemis
10. Mounikhion (Μουνιχιών), first month of spring, sacred to Artemis
11. Thergelion (Θαργηλιών), second month of spring, sacred to Artemis and Apollo
12. Skirophorion (Σκιροφοριών), third month of spring, sacred to Athena

Each month had approximately 30 days (more on that “approximately” part in a bit), divided into three periods of ten days each (which we’ll call “decades”):

 Moon waxing Moon full Moon waning New Moon 11th later 10th 2nd rising 12th 9th waning 3rd rising 13th 8th waning 4th rising 14th 7th waning 5th rising 15th 6th waning 6th rising 16th 5th waning 7th rising 17th 4th waning 8th rising 18th 3rd waning 9th rising 19th 2nd waning 10th rising earlier 10th Old and New

The first day of the month was officially called the New Moon, or in Greek, the νουμηνια, the date when the Moon would officially be sighted on its own just after syzygy.  The last day of the month was called the Old and New, or ενη και νεα, which was the actual date of the syzygy between the Earth, Moon, and Sun.  The last day of the second decade and the first of the third decade were both called “the 10th”, with the earlier 10th being the first day and the later 10th being the second.  Days in the months would be referred to as something like “the third day of Thargelion waning”, or Thargelion 28.  Only days 2 through 10 were referred to as “rising”, and days 21 through 29 were referred to as “waning”; the middle block of days from 11 to 19 were unambiguous.  When a month was “hollow”, or had only 29 days instead of 30, the 2nd waning day was omitted, leading to the 3rd waning day becoming the penultimate day of the month instead of the 2nd waning day.  Since this was all based on observation, there was no hard and fast rule to determine which month was hollow or full without the use of an almanac or ephemeris.

At this point, we have enough information to start applying the Greek alphabet to the days.  As mentioned before, there are fewer letters in the Greek alphabet than there are days, so there are some days that are simply going to remain letterless; like the intercalary days of the solar calendar, these might be considered highly unfortunate or “between” times, good for little except when you have a sincere need for that bizarre state of day.  A naive approach might be to allot the 24 letters of the Greek alphabet to the first 24 days of the lunar month, then leave the last six or seven days unallocated, but I have a better idea.  If we include the otherwise useless obsolete letters digamma (Ϝ), qoppa (Ϙ), and sampi (Ϡ), we end up with 27 days, which is 9 × 3.  In using the Greek letters as numerals (e.g. isopsephy), letters Α through Θ represent 1 through 9, Ι through Ϙ represent 10 through 90, and Ρ through Ϡ represent 100 through 900.  In other words,

 Α/1 Β/2 Γ/3 Δ/4 Ε/5 Ϝ/6 Ζ/7 Η/8 Θ/9 Ι/10 Κ/20 Λ/30 Μ/40 Ν/50 Ξ/60 Ο/70 Π/80 Ϙ/90 Ρ/100 Σ/200 Τ/300 Υ/400 Φ/500 Χ/600 Ψ/700 Ω/800 Ϡ/900

In this system of numerics, it’s easy to group the letters into three groups of nine based on their magnitude.  This matches up more or less well with the three decades used in a lunar month, so I propose giving the first nine letters to days 1 through 9 (Α through Θ) and skipping the 10th rising day, the second nine letters (Ι through Ϙ) to days 11 through 19 and skipping the earlier 10th day, and the third nine letters (Ρ through Ϡ to days 21 through 29, and leaving the Old and New day unassigned.  If the month is hollow and there is no 2nd waning day for Ϡ, then the Old and New day (last day of the month) is assigned Ϡ.  Letterless days might repeat the preceding letter; thus, the 10th day of the month (or the 10th rising day) might be called “second Θ”, but still be considered effectively letterless.

With the usual Attic festivals celebrated monthly (they treated the birthdays of the gods as monthly occurrences), the lunar month with all its information would look like the following:

 Day Name Letter Festival 1 New Moon Α Noumenia 2 2nd rising Β Agathos Daimon 3 3rd rising Γ Athena 4 4th rising Δ Heracles, Hermes, Aphrodite, Eros 5 5th rising Ε 6 6th rising Ϝ Artemis 7 7th rising Ζ Apollo 8 8th rising Η Poseidon, Theseus 9 9th rising Θ 10 10th rising — 11 11th Ι 12 12th Κ 13 13th Λ 14 14th Μ 15 15th Ν 16 16th Ξ Full Moon 17 17th Ο 18 18th Π 19 19th Ϙ 20 earlier 10th — 21 later 10th Ρ 22 9th waning Σ 23 8th waning Τ 24 7th waning Υ 25 6th waning Φ 26 5th waning Χ 27 4th waning Ψ 28 3rd waning Ω 29 2nd waning Ϡ Omitted in hollow months 30 Old and New — (Ϡ if hollow month)

That’s it, really.  All in all, it’s a pretty simple system, if we just take the lunar months as they are, and is a lot easier than the complicated mess that was the SGC.  Then again, that’s no fun, so let’s add more to it.  After all, the fact that the months themselves are 12 and the Greek letters are 24 in number is quite appealing, wouldn’t you say?  And we did add letters to the months in the SGC, after all, so why not here?  We can also associate the months themselves with the Greek letters for grammatomantic purposes; if we assign Α to the first month of the year, we can easily get a two-year cycle, where each of the months alternates between one of two values.  For example, if in one year Hekatombaion (first month of the year) is given to Α, then by following the pattern Skirophorion (last month of the year) is given to Μ; Hekatombaion in the next year is given to Ν to continue the cycle, as is Skirophorion in the next year given to Ω.  The next Hekatombaion is given to Α again, and the cycle continues.  Note that the obsolete Greek letters digamma, qoppa, and sampi would not be used here; I only used them in the lunar month to keep the days regular and aligned properly with the decades.

The thing about this is that the lunar months don’t match up with the solar year very well.  Twelve lunar months add up to about 354 days, and given that a solar year is about 365 days, the year is going to keep drifting back unless we add in an extra intercalary (or, more properly here, “embolismic”) month every so often to keep the calendar from drifting too far.  Much as in the SGC with the intercalary days, we might simply leave the embolismic month unlettered in order to keep the cycle regular.  Days within this month would be lettered and celebrated as normal, but the month itself would be otherwise uncelebrated.  For the LGC, we would add the embolismic month at the end of the year, after Skirophorion, so that the next Hekatombaion could occur after the summer solstice as it should.  I depart from the Athenian practice here a bit, where other months would simply be repeated (usually Poseideon).

Of course, figuring out which years need the embolismic month is another problem.  To keep the cycle regular, we’d need to add in an embolismic month one year out of every two or three.  Although there’s no evidence that the Athenians used it, I propose we make use of the Metonic cycle, a period of 19 years in which 12 of the years are “short” (consisting of only 12 months) and 7 are “long” or leap years (consisting of 13, or 12 months plus an embolismic month).  This cycle has been in use for quite some time now in other calendrical systems, so let’s borrow their tradition of having years 3, 6, 8, 11, 14, 17, and 19 be long years, and the other years being short.  Just as with the months, the 12 short years might be assigned letters of their own, while the long years would be unlettered due to their oddness (in multiple senses of the word).  Since the Metonic cycle has an odd count of years, two of these cycles (or 38 years) would repeat both a cycle of letter-years as well as letter-months in the LGC.  Since the use of an epoch for the LGC isn’t as necessary as in the SGC, figuring out where we are in the current Metonic cycle can be determined by looking at another calendar that uses it; I propose the Hebrew calendar, which does this very thing.  In this case, the most recent Metonic cycle began in 1998, with the long years being 2000, 2003, 2005, 2008, 2011, 2014, and 2016; the next Metonic cycle begins in 2017.  The two Metonic cycles, which we might call a LGC age or era,  starting in 1998 and ending in 2035, are below, and the same cycle is repeated forward and backward in time for every 38 years.

 Year Cycle Length Letter 1 2 3 4 5 6 7 8 9 10 11 12 (13) 1998 1 12 Α Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ 1999 2 12 Β Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω 2000 3 13 — Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ — 2001 4 12 Γ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω 2002 5 12 Δ Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ 2003 6 13 — Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω — 2004 7 12 Ε Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ 2005 8 13 — Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω — 2006 9 12 Ζ Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ 2007 10 12 Η Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω 2008 11 13 — Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ — 2009 12 12 Θ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω 2010 13 12 Ι Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ 2011 14 13 — Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω — 2012 15 12 Κ Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ 2013 16 12 Λ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω 2014 17 13 — Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ — 2015 18 12 Μ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω 2016 19 13 — Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ — 2017 1 (20) 12 Ν Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω 2018 2 (21) 12 Ξ Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ 2019 3 (22) 13 — Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω — 2020 4 (23) 12 Ο Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ 2021 5 (24) 12 Π Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω 2022 6 (25) 13 — Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ — 2023 7 (26) 12 Ρ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω 2024 8 (27) 13 — Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ — 2025 9 (28) 12 Σ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω 2026 10 (29) 12 Τ Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ 2027 11 (30) 13 — Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω — 2028 12 (31) 12 Υ Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ 2029 13 (32) 12 Φ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω 2030 14 (33) 13 — Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ — 2031 15 (34) 12 Χ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω 2032 16 (35) 12 Ψ Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ 2033 17 (36) 13 — Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω — 2034 18 (37) 12 Ω Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ 2035 19 (38) 13 — Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω —

A few others of these cycle-epochs include the following years, covering the 20th and 21st centuries, each one 38 years apart from the previous or next one:

• 1884
• 1922
• 1960
• 1998
• 2036
• 2074
• 2112

Creating an epoch to measure years from, although generally useful, isn’t particularly needed for this calendar.  After all, the Attic calendar upon which the LGC is based was used to determine yearly and monthly festivals, and years were noted by saying something like “the Nth year when so-and-so was archon”.  Similarly, we might refer to 2013 as “the 16th year of the 1998-age” or 2033 as “the 35th year after 1998”.  In practice, we might do something similar such as “the sixth year when Clinton was president” or “the tenth year after Hurricane Sandy”; measuring years in this method would still be able to use the system of letter-years in the LGC, simply by shifting the start of the epoch to that year and starting with letter-year Α.  The Metonic cycle would continue from that epoch cyclically until a new significant event was chosen, such as the election of a new president, the proclamation of a peace between nations, and so forth.

Associating the letters with the years and months here is less for notation and more for divination, since the LGC is an augmentation of the Attic festival calendar (with some innovations), and not a wholly new system which needs its own notation.  That said, we can still use the letters to note the years and the months; for instance, the 16th year of the cycle given above might be called the “year Λ in the 1998-age”, while the 17th year (which has no letter associated with it) might be called just “the 17th year” or, more in line with actual Attic practice, “the second Λ year”, assuming that (for notational purposes) a letterless year repeats the previous year’s letter.  Likewise, for embolismic months, we might say that the 12th month of a year is either “the Μ month” or “the Ω month”, and the 13th month of a year (if any) could be said as “the 13th month”, “the empty month”, or “the second Μ/Ω month” (depending on whether the preceding month was given to Μ or Ω).

Converting a date between a Gregorian calendar date and a LGC date or vice versa is much easier than the SGC conversion, but mostly because it involves looking things up.  To convert between a Gregorian calendar date and a LGC date:

1. Find the year in the cycle of the LGC ages to find out whether the year is a long or short year.
2. Count how many new moons have occurred since the most recent summer solstice.
3. Find the date of the current moon phase.

For instance, consider the recent date September 1, 2013.  This is the 16th year in the LGC age cycle, which has only 12 months and is associated with the letter Λ.  The summer solstice occurred on June 21 this year, and the next new moon was July 8, marking the first month of the LGC year.  September 1 occurs in the second month of Metageitnion, associated with the letter Ξ this year which starting on the new moon of August 7, on the 26th day of the lunar month, or the 5th waning day, associated with the letter Χ.  All told, we would say that this is the “fifth day of Metageitnion waning in the year Λ of the cycle starting in 1998”; the letters for this day are Λ (year), Ξ (month), and Χ (day).

Now that your brain is probably fried from all the tables and quasi-neo-Hellenic computus, we’ll leave the actual uses of the LGC for the next post.  Although the uses of the SGC and LGC are similar in some respects, the LGC has interesting properties that make it especially suited for magical work beyond the daily divination given by the letter-days.  Stay tuned!

# Thoughts on a Grammatomantic Calendar

Earlier this year, I produced my first ebook, a short text detailing the history and use of grammatomancy, or divination using the Greek alphabet much as one might use runes for divination. It’s an interesting system, and I combined the ancient oracular meanings of the letters with their isopsephic (gematria) meanings, stoicheic (planetary/elementary/astral) meanings, and qabbalistic symbolism to produce a full divination system suitable for any student of the magical arts. It got real complicated real fast, but also real complete in the process. (If you don’t have a copy, stop being lazy and get one here.)

As some of my readers may know, I make use of this every day (mostly) for my Twitter/Facebook feeds under the posts “Daily Grammatomancy”. It’s helped me and others plan our days out, using a simple oracle for how the day will go; the question I ask for our mutual and communal benefit is “for myself and for all who come in contact with my words, for this day, this very day: how best should we live our lives in accordance with the divine will of the immortal gods?”. For some people, it’s no better than a newspaper horoscope; for others, it hits dead on time and time again.

Doing this for a while has lead some of my friends to start pursuing their own daily divination methods. One such friend, Raven Orthaevelve (who is a fantastic artist and crafter whom you should totally buy and commission things from for anything fancy, magical, or otherwise), has started using the Mayan calendar as a divination tool. This isn’t any 2012 bullshit, either; the Mayan calendar was known for being a reasonably complex set of interlocking cycles. One such calendar used for these cycles is the tzolk’in, a 260-day calendar made up of 13 20-day “months”. Each day has a particular name and divinatory meaning which forms the basis of much of Mayan divination, natal astrology, and prognostication. Raven posts her interpretations of the tzolk’in daily on her Facebook, and will eventually build in other Mayan cycles into the mix for a more complex and complete daily prognostication.

In some sort of weird feedback loop, this has started to help me pursue my own idea of a cyclical divination using Greek letters. In other words, although the daily grammatomantic divination would be helpful for specific days, a day might generally have a particular meaning based on its location in a cycle of days; combining the two can help focus knowledge and energy for particular problems, much as one might combine the cycle of planetary days and planetary hours for rituals. Interesting as this idea might be, though, it’d be incredibly difficult; there’s no information I can find that this was done in ancient times, so it’d be a new innovation. Add to it, the development of any kind of calendrical cycle is difficult (as my experiments with forming a ritual calendar for planning things out have shown).

One option I might explore is just using the 24 letters of the Greek alphabet in making a type of calendar. 15 repetitions of this cycle would produce 360, a very nice number indeed; if I were to tie it to the solar year, then it’d produce 5 or 6 intercalary days that would not be associated with any particular letter. This kind of practice isn’t uncommon by any means: the Mayans used a similar practice with their haab’ calendar: 18 cycles of 20 days, again producing 360 days with five days (wayeb’) at the end; the ancient Egyptians and modern Copts use 12 cycles of 30 days, again 360 days, plus five or six days at the end. 24 is a pretty convenient number, I have to admit, especially with its divisors and amenability to larger cycles of 12, 360, and the like.

Plus, if I were to use this cycle of 15 months of 24 days, I could further associate each month with a particular letter, which could afford another general cycle around the day-letter cycle. Say that we associate the first month of the first year with Α, then the second with Β, and so forth. Because the year only has 15 months, the last month of the first year would be Ο, and the first month of the second year would be Π; likewise, the first month of the third year would be Η, the first month of the fourth year would be Χ, and so forth. This produces a cycle such that every 9th year has the first month starting with Α; thus, there are eight distinct years with this month-letter cycle. And if we have a month-letter cycle, we could also expand this to a year-letter cycle, such that three such month-letter cycles form one year-letter cycle, or 24 years (8 × 3 = 24). Alternatively, we might have a great year-letter cycle, where one month-letter cycle is given a letter, and 24 month-letter cycles completes a great year-letter cycle; this would be a cycle of 8 × 24 = 192 years. At this point, I’d just be forming cycles for the sake of cycles for a kind of neo-Greek long count calendar, but it’d be nifty all the same for finer, long-term gradations of influences.

To use such a cycle, however, I’d need to use a particular day as a start date, at least for the months. Although the Greek alphabet oracle as I use it was found in modern Turkey, different parts of ancient Greece used different calendars with different start dates for individual years. The Attic calendar, about which we know the most among all ancient Greek calendars, started on the first new moon after the summer solstice; other Greek calendars often started off between autumn and winter. For simplicity, I’d say that either the spring equinox (to tie it in with astrology) or the summer solstice (to tie it in with Athenian practice) would be the official start date. Thus, the five or six days leading up to this start date (according to the Gregorian calendar) would be the intercalary days, which would have no letters assigned to it; alternatively, I might devise a scheme to associate particular letters with these intercalary days based on specific properties of the letters. This doesn’t even mention where the month-cycles (years of day-letters) would begin, or what the anchor date might be.

If I were to use the Attic practice of using the summer solstice as the start date, though, why not actually go ahead and use the Attic calendar itself as the basis for my cycle? The Attic calendar, specifically the festival calendar used to determine festivals and rituals, was a lunisolar calendar. There were 12 months as determined by the observation of the Moon; months began on the first sighting of the New Moon just after syzygy (νουμηνια, “new moon”), and ended on the day of the syzygy itself (ενη και νεα, “old and new”). Thus, the months could be 30 days (full months) or 29 days (hollow months). A month was divided into three periods of ten days each, which I’ll call decamera; if the month was hollow, then the third decameron would have only nine days, with the usual 29th day being omitted entirely.

The problem with using this type of lunisolar calendar is that there are more days in a month (29 or 30) than there are letters in the Greek alphabet (24). Even if I were to include the obsolete letters digamma, qoppa, and sampi for a total of 27 letters, this would still leave three days leftover. This might be remedied by throwing “letterless” days into the mix, on which no advice can be given, as well as making the obsolete letters effectively “letterless” since they have no associated oracles or stoicheic meaning. They have isopsephic meaning, however, which can be substituted with Hebrew gematraic meanings (and, through them, Hebrew stoicheic meanings), but this is starting to overreach and combine different traditions.

However, the use of those three decamera within each month does lend itself well to the Greek alphabet, assuming we use the full body of 27 letters. Using isopsephy, the letters Α through Θ (including digamma) are given to values 1 through 9, Ι through qoppa are given to 10 through 90, and Ρ through sampi are given to 100 through 900. Thus, we have nine letters per decameron, of which two days per period are letterless (or one day for the final period in the case of a hollow month); one for the obsolete letter in the mix and one extra letterless day at the end of the period. In this manner, we’d have a method to create a grammatomantic lunisolar calendar, which would be interesting to use. There’d be gaps in the calendar, of course, but it’s no worse than other magical calendars I’ve seen, e.g. PGM VII.155-167 or the Munich Manual, for determining which day of a given month is good or bad for magic or divination.

Using a year of 12 months is convenient, and can make the process of assigning letters to each month much simpler: a month-cycle of two years can be had here, since two years of 12 months produces 24 months, one for each letter. That said, the issue with lunisolar calendars is that the months get out of sync without an embolismic month, or intercalary month every so often. Using the ancient Metonic cycle of 19 years, there would be 12 “short” years (years with 12 months) and 7 “long” years (years with 13 months). The embolismic month could be held as letterless and placed at the end of the year in long years. Thus, every two years would complete one month-letter cycle in this lunisolar scheme; due to the parity of the Metonic cycle, every 38 years would complete one year-letter cycle. Using the Babylonian and Hebrew method of assigning embolismic months, years 3, 6, 8, 11, 14, 17, and 19 would be long years.

So, that turned out to be a much longer discussion on calendrical cycles for divination than I intended. Then again, calendars and cycles have never been easy to work with for any culture or era. In all honesty, the use of the simple 15 months of 24 days plus intercalary days is highly appealing for the sake of its simplicity and ability to lend itself to cycles within cycles. There is something to be said for the attribution of letters to the lunisolar months, though, especially for the sake of timing rituals or determining favorable lunar influences for a given letter. I’ll try drafting the rules and algorithms for these two types of grammatomantic calendars, along with date calculation methods, and see where that gets me.  The next few posts will go over these two types of calendars, one based on the solar cycle of the seasons and one based on the lunisolar cycle of the lunar months as tied to the seasons, so stay tuned!