Yes, yes, I am still working on my geomancy textbook!  As I’ve said before, it’s a long project, and by necessity it’s not the number one priority in my life; between a full-time job, heavy involvement in my religious community, managing several kinds of online presence, and my own routines and practices, working on my book is definitely a priority but not the priority.  If I had days empty of all other tasks, it’d be a different story, but here we are.  Besides, the book has been in progress since 2013, back at a point where I now think I was wholly unqualified to write such a book.  (I still think I am unqualified to write such a book, not least because I’ve made a number of discoveries, innovations, and corrections to what I knew earlier, but here we are.)

One of the fun parts of the book for me to write is the postscript.  It’s an appendix that, rather than focusing on the meat-and-bones of geomantic techniques and practices, I talk a small bit about my own thoughts, views, and opinions on certain techniques and how my own practice prioritizes certain techniques over others, or my value-estimates of certain techniques.  After all, though there are hundreds of different techniques that one can use in geomantic divination, in any given chart I might only use a handful of them, some I use generally for every reading and others I bust out for particular situations.  Almost all the techniques have some value, but some have more value than others.  I talk a bit about what I think of such things in the postscript as a kind of final letting-my-hair-down moment, where I get to drop a little of the academic and technical style I use throughout the book and get a little personal in my practice.

The postscript really isn’t a place for me to introduce or talk about any particular techniques at length, though—except one: methods to determine names or letters with geomancy.  As I’ve mentioned before on my blog, the methods to determine names is something that would be sorely useful for geomancers, and a number of historical authors mention methods to do so, most of all Christopher Cattan who introduces several “rules” for associating the figures with letters and a number of methods to use them.  John Heydon, likewise, introduces several such sets of associations for different scripts, but largely references the same methods Cattan uses.  John Michael Greer, continuing the vein of carrying on such information especially as it was republished over and over again in the late Renaissance, gives a similar set of attributions in his “Art and Practice of Geomancy”.

It’s all a shame, though, because I’ve never gotten these methods to work.  In my past experiments with them, I kept getting garbage answers with chance results.  Quoth my earlier article:

Alas, however, I have to consign a geomantic technique to the failure pile, and it’s not for lack of trying: determining names.  While it would make sense conceptually that one could determine names with geomancy, I have never been able to get such name charts to work right, from the first time I ran a name chart years ago up until the present day.  Add to it, I’ve found several methods to determine names with geomancy, and several ways to associate the letters to the figures, and I’ve tried them all, none of them giving anything remotely resembling an accurate answer.  This frustrates me to no end, because why the hell would this one technique not work when nearly every other technique I’ve tried has given me useful results?  This is especially frustrating, since being able to predict names would be exceptionally useful in the world, from determining the names of cities one might be successful in to determining the names of future spouses. …

But even using any of the techniques with any set of correspondences, I kept coming up with wrong answers.  If I were lucky, some of the letters in the actual name I was trying to find might appear at random places in the chart, but this was by no means guaranteed.  I did notice a slight tendency for some of the letters to appear in houses II, V, and VIII, but there was no pattern for which letters (start, medial, end) appeared within them.  I even tried using the values of the Greek, Hebrew, and Celestial Hebrew associations that Heydon gives (untrustworthy as his stuff tends to be) to see if it would get me anything closer than the Roman script association; nada.  Plus, many of the techniques assumes there to be at least four letters or syllables in a name; many names I ended up asking about after I did a reading on them had one or two syllables, or had even just three letters, and these techniques don’t specify what to do in the case of really short names.

It seems, also, that I’m not the first person to complain about these methods, not by far.  In addition to my own colleagues and contacts in the present day who largely give the same conclusions I have, the French geomancer Henri de Pisis gives in his 17th century book Opus Geomantiae Completum in libros tres divisum (reproduced as part of Fludd’s later work Fasciculus Geomanticus) gives the following complaint when he introduces these methods (translation mine from Latin):

So as to know someone’s name. I might have put this and another table of the same from Cattan, yet given how useless and hollow it is, I freely suppress it, lest it impose onto this very art which usually predicts with certainty. By this understanding, I would have omitted it and the following chapter, as with things uncertain and generally wrong, if not for that we would see what even a single author maintains …

In truth, it has always escaped me as to the use they make of these numbers here, for nobody thus far has been able to discover their reasoning; neither Gerard of Cremona, nor Geber, nor Pietro d’Abano, nor myself, nor any others besides Cocles and Cattan have discerned the reasoning of the numbers or of the letters of names. It can essentially be seen that Cattan and Cocles would have relaxed this art to such a freewheeling extent into the form of some game, such as the casting of dice or dominoes, for the troublesome cheating of long nights or for the future coaxing of a droll joke, and a good many use it for this and will have had nothing certain placed in the art. In other words, since they are unaware of that which is superfluous to the art, they are unestablished in the foundations of this very art, and are only outsiders into contempt of it. I suggest that these methods be rejected.

It’s frustrating, especially for someone like de Pisis to have written so bluntly about this in a way he doesn’t elsewhere in Opus Geomantiae; he only includes these methods because others have written about them, and that only bitterly and begrudgingly.  This is all the more frustrating because Arabic geomancers make claims to predict names and letters as a matter of course, though because I speak neither Arabic nor Urdu nor Farsi, it’s hard for me to find what methods they use to validate it and see whether they can walk the talk or if they’re just full of hot air.

Now, skip ahead a few years.  The Geomantic Study-Group on Facebook is thriving with over a thousand members, including a good number from Arabic-speaking countries who are, God bless them, actually willing to share and discuss Arabic methods of geomancy.  One of them even goes so far as to include a list of those fancy apparati of Arabic geomancy, taskins, though I prefer an alternate term for them now, dā`ira (plural dawā`ir), which is commonly found in Urdu and Farsi texts, and which literally mean “cycle”.  These things are fascinating for Western geomancers to look at, because we have no parallel for them; they’re a combination of correspondence as well as technique unto themselves, enforcing particular orders of figures for different needs.  Depending on the tradition of Arabic geomancy you’re looking at, some geomancers claim that there are 16 cycles, others 28, or even as many as 400 or more, some kept secret for mystical and magical ends.  Some dawā`ir are clearly organized along mathematical or otherwise clearly understood principles, such as the dā`ira-e-abdaḥ which organizes the figures according to their binary numeral meanings (reading Laetitia as 1000 as 1, Rubeus as 0100 as 2, Fortuna Minor as 1100 as 3, and so forth); others are far more obscure as to why certain figures are arranged in certain ways.

So this list of dawā`ir is shared in the group, and happily the poster who shared it cited a particular academic: Dr. Matthew Melvin-Koushki, currently of the University of South Carolina, one of whose research interests is the occult sciences in Islam.  In his paper “Persianate Geomancy from Ṭūsī to the Millennium: A Preliminary Survey” (in Nader El-Bizri and EvaOrthmann, eds., Occult Sciences in Pre-modern Islamic Cultures, Beirut: Orient-Institut Beirut, 2018, pp. 151-99), Melvin-Koushki lists seven such cycles:

The various regional schools of geomantic thought are therefore to be distinguished by the ‘cycles’ (sg. dāʾira) they prefer to employ. A cycle, Hidāyat Allāh explains, is simply a specific sequence (tartīb) of the 16 geomantic figures, with each sequence featuring different types of correspondences — elemental, humoral, temporal, astrological, lettrist, etc. And the four cycles he identifies above are far from the only ones in common use. The geomancer has at his disposal a rather larger number of cycles; which he employs in any given reading depends on the nature of the information being sought and the degree of detail required. Hidāyat Allāh lists seven by way of example:

1. The Occupation (sakan) cycle is the most basic; it begins with Laetitia and ends with Via.
2. The Constitution (mizāj) cycle tells the querent what day he can expect to realize his desire; it begins with Acquisitio and ends with Cauda Draconis.
3. The BZDḤ cycle, aka the Number (ʿadad) cycle, is used to tell periods of time; it begins with Puer and ends with Populus.
4. The Letter (ḥarf) cycle is used to reveal names (a very popular application of the science); it too begins with Laetitia and ends with Via, with the first twelve figures being assigned two letters each and the last four figures only one.
5. The Arabic Alphabet (abjad-i ʿarabī) cycle, aka the Element (ʿunṣur) or Interior (dākhil) cycle; it begins with Laetitia and ends with Populus.
6. The ABDḤ cycle, aka the Element (ʿunṣur) or Arabic Alphabet (abjad-i ʿarabī) cycle, which is also popular; it too begins with Laetitia and ends with Populus.
7. The Most Complete (aṣaḥḥ) cycle, which has a different organizing principle and procedure; it begins with Laetitia and ends with Via.

Note that fourth cycle Melvin-Koushki lists: the ḥarf cycle, the word literally meaning “letter” (as in characters of an alphabet).  This cycle goes in the following order:

1. Laetitia
2. Tristitia
3. Rubeus
4. Albus
5. Fortuna Maior
6. Fortuna Minor
7. Caput Draconis
8. Cauda Draconis
9. Puer
10. Puella
11. Acquisitio
12. Amissio
13. Populus
14. Carcer
15. Coniunctio
16. Via

Melvin-Koushki says that the first twelve figures (Laetitia through Amissio) get two letters each, and the final four (Populus through Via) get one letter each.  Looking through contemporary texts on Arabic geomancy (despite my lack of knowledge of Arabic/Farsi/Urdu, I can still pick out patterns and particular words well enough to find them!), we get the following correspondences of figures to letters:

Note the order of how the letters go, first down the left column then down the right: this is the traditional abjadī order of the Arabic script, the same one in use for all other Phoenician-derived scripts like Greek and Hebrew.  The fact that the last four figures in the ḥarf cycle have only one letter each are also the liminal figures that are neither entering nor exiting might be because these four figures are special.  More realistically, though, it’s because there are 28 letters in the Arabic script, which means that some figures would get two letters and others only one; because there are 16 figures, 16 × 2 = 32, and 32 – 28 = 4.  If you just start assigning the letters one by one to the figures, you’d run out for the last four.  This raises the question, which came first, the order of the figures, or the ordering of the letters to which the figures were then mapped?  It’s unclear which came first to me, but we can pick out some interesting structural notes about the ḥarf cycle:

• The first 12 figures are given in reversion pairs: Laetitia/Tristitia, Rubeus/Albus, etc.
• The first four figures are the “pure elemental” figures, each with seven points.
• The last four figures are all liminal figures, each of which is their own reversion; the first two are considered the stable liminal figures, the latter two the mobile liminal figures, progressively going from the most stable to the most mobile.

I also want to note that the source Melvin-Koushki is referencing came from the late 16th century, and his sources likely came from much older ones; by that point, geomancy was already around 600 or 700 years old.  Regardless, this cycle is still found in many works even today as a means to predict names.  (I have also seen the ABDḤ/binary-numeral cycle used for this same purpose, but it seems like that’s less popular of a choice than using this specific cycle, though the mechanism is the same.  I don’t know how common using the ABDḤ cycle is for this purpose, or where it might be centralized.)  Although I haven’t yet found much in English or another language I know yet about how to specifically use this cycle for divining names, at least I know how they associate the figures with letters, which is pretty neat unto itself.

I bring this up because, while going over my draft for my postscript in my book, I returned to that section about how Western geomancy has methods for determining names.  I originally wrote the seed for that section in the aforementioned blog post of mine back in 2014, and I basically copied the same tables (in a more intelligible way and broken down by author or source) into my book.  While I was revising that particular section, something about the order of how Cattan, Heydon, and Case associated the figures to the letters…something about it struck me as familiar.  I normally use the planetary order of the figures in my posts and tables (lunar figures, Mercurial figures, Venereal figures, …, nodal figures), but it struck me that several Western authors all had it that Laetitia was given to A, Tristitia to B, Rubeus to C, and so on.  They don’t all agree with each other in some of the associations, and Cattan and Heydon have other rules that give other letters to the figures, but it’s clear they were all drawing on the same source in one form or another, and…hm.  Neither the similarities between them nor that same order could be given to chance.

Digging out my ancient binder of geomancy notes from when I was in college, I got out my transcript of one of the earliest Western works on geomancy, Martin of Spain’s work “De Geomancia”, written sometime in the 1200s.  Dr. Laurel Means has a version of it in Popular and Practical Science of Medieval England (Lister M. Matheson, ed., Michigan State University Press, 1994), and I was able to get a text transcript of it while in college, though I’ve since lost the original source and the transcript file I was working on, though I did save a copy.  I remembered this because it has an early association of the figures with letters from well before Cattan or the others, and I wanted to see how it’d match up.  Surprise: it did, more than I expected, even if I’m missing associations for two of the figures.  Though Martin of Spain gives anywhere from one to five letters to the figures, the first of them typically matches with the expected one and seems to be the “primary” letter.  All these Western sources all seemed to share the same basic order of the figures, starting with Laetitia and Tristitia and continuing from there.  There are some variations, but it’s all fundamentally the same thing.

To compare what I’m seeing, here’s a table that associates the letters of the alphabet with the figures from Christopher Cattan (specifically his First Rule), John Heydon (the “First Rule” for English, with annotations), John Case, and Martin of Spain (more below because this is weird):

Admittedly, Martin of Spain’s attributions are a little weird; he gives a set of letters for each figure, roughly in alphabetical order per figure, so a bit of sussing needs figuring out; additionally, the letters “l” and “y” are not allocated anywhere, but there is an allocation for the obsolete letter yogh (ʒ), which I interpreted as “y” above.  The full set of associations from Martin of Spain are:

Anyway, none of the associations we have in Martin of Spain, Cattan, Heydon, or Case give a figure for the letter Z, and the letters I/J and U/V/W weren’t classified as separate letters until recently, anyway.  As always, Case agrees with Heydon, as I’m pretty sure Case’s Angelical Guide was based on Heydon’s Theomagia, and both differ from Cattan in some minor ways.  Martin of Spain’s order starts off clear, but the order gets really mixed up and unclear towards the end.  Cattan’s order seems to be the most orderly, and preserves almost but frustratingly not quite the same order as the ḥarf cycle from before, with the following changes:

• Cattan has Fortuna Minor and Fortuna Maior in positions 5 and 6; the ḥarf cycle switches these
• Cattan has Puella and Puer in positions 9 and 10; the ḥarf cycle switches these
• Cattan has Via, Carcer, Populus, and Coniunctio as the final four figures; the ḥarf cycle has Populus, Carcer, Coniunctio, and Via

What we’re arriving at is that the Western rules for assigning letters to the figures is clearly a continuation of the same cycle associations that began in the Arabic tradition, even from a very early period in Western geomancy, which indicates that the ḥarf cycle definitely dates back to the late 1200s, probably earlier, making it a very early arrangement of figures, indeed.  At least in the western part of Europe (i.e. Spain as opposed to Greece), this was likely brought in at an early point along with the rest of geomantic technique, and held on in some form or another by a handful of geomancers.  It’s unclear to me exactly how popular this method or association was, since I haven’t found more than a handful of resources that give such an association and most of them tend to be the larger works on geomancy that date from Cattan and onwards, but it may well be that this system was held onto, perhaps with some corruptions or changes, which would explain the small changes in Cattan’s order versus the ḥarf cycle.

The other major difference is how the letters get assigned to the figures in their cycle.  Rather than how the Arabic method goes through the cycle of figures and gives each figure one letter in turn, which results in a bunch of figures at the start with two letters and a few at the end with only one, the European method seems to almost be reverse: double up the letters at the end of the cycle and work forward until the rest of the figures at the start have only one letter each.  Given how straightforward the association method would be, I’m not sure how the method changed so drastically; either several corruptions happened along the way, or someone innovated a variation on the system.  I can’t seem to trace sources back past Cattan, or at least find any in an accessible form, so it’s unclear whether Cattan or his predecessors carried on the same tradition that Martin of Spain wrote about, whether his method came from another variant closely related to it, or whether he reimported an Arabic method and customized it for European needs.

What would it look like if we were to use the ḥarf cycle order of the figures and the same method, but applied it to the Roman script?  Considering that the Roman script that we use nowadays has shifted a bit from Renaissance usage, notably with the introduction of a few more letters (J from I, V and W from U, and Z), we can envision two versions of this, a “Renaissance Roman” ḥarf association of the figures with the letters as it was done in the days of Cattan et al. with 23 letters, and a “Modern English” association that uses all 26 letters of the modern English alphabet but done in the same way.  Below is what we would get from using those methods, alongside Cattan’s association for comparison (with the mis-ordered letters, e.g. Fortuna Maior and Fortuna Minor, in bold italic).  Heck, we can even come up with a Cattan-style association of the letters, using the ḥarf ordering (to fix the irregularities we might have seen from before) but using the same Western-style doubling-up of successive letters at the end:

In this light, let’s point out two things about Cattan’s original style (which I’m taking as the default Western letter association rule, which was an earlier version of what Heydon and Case later used):

• Really, why is there no Z in the Renaissance Roman scheme, or even Cattan’s original scheme?  As I mentioned earlier, Z was barely considered a letter in English until comparatively recently, so it’s not completely surprising that medieval, Renaissance, and even early modern texts on geomancy would omit it from such an association scheme.  Yet, French (for that matter, many forms of Romance languages) definitely uses the letter Z in its language regularly, so it’s odd that French or Italian would omit this letter.  Note how it would fall in the ḥarf-style Renaissance Roman scheme, as a letter corresponding to Caput Draconis.  This, however, would give its reverse figure Cauda Draconis no corresponding double letter, because the Roman script including Z would have 23 letters, and an odd number would mean one of these reversion-pairs would go unassigned.  So, this letter would have to be omitted to keep the system clean, and would probably logically be merged with S (as part of Carcer).
• Even then, why does the Cattan scheme double up successive letters at the end, rather than allocate sequential letters cyclicly through the alphabet?  It might be more for a superficial resemblance or mirroring of how the ḥarf cycle associations work for Arabic, where the final positions are given to the liminal figures which were seen as “breaking the pattern” in some special way.  Because 22 letters get nowhere near those final four figures (as the ḥarf-style Renaissance Roman scheme shows) and because we might still want to make those final figures special in some way, the doubling-up of successive letters at the end could be seen as a compromise to keep the final few figures special while still allocating the letters to the figures in an orderly way.  It’s a major departure from the logic of the ḥarf cycle method, but it’s a method all the same.

So, let’s say that we have our pick now of these four systems.  Which would I recommend to use?  Given what the original ḥarf cycle logic was, I would throw my hat in for the ḥarf-style modern English associations above, but that’s also because I use the English language, and though the Renaissance Roman script is just an earlier version of the English alphabet, I see no reason to use an outdated orthography that omits several important letters that have not been considered allographs or variants of others for several hundred years now.  The same method of straightforwardly allocating the letters of one’s writing system in order to the geomantic figures in the ḥarf cycle can be used for any alphabetic or abjadic script.

Even with this, there are still several important questions that are still left unanswered:

• It’s clear that alphabets or abjads that have an even number of letters would be favored, because it keeps the reversal pairs intact, so that each figure in the pair has the same number of letters.  What about scripts with an odd number of letters?  Does it really matter that much to keep reversal pairs intact?
• Why are the figures in the ḥarf cycle placed in this order at all?  Is there an organizing principle behind it, or was it more inspired than devised?
• Did the ḥarf cycle come first and then the association with the letters, or did the idea of divvying up the 28 letters of the Arabic script come first and the figures associated with those letters afterwards?  If the latter, it could explain why the four liminal figures just so happen to be at the end of the cycle where they get one figure each.  But even then, why would the pure element figures Laetitia, Tristitia, etc. be at the front in that order?  Reading the figures as elements, they could be read as Fire-Earth-Air-Water (my modern system or just using the points of those elemental lines) or as Air-Earth-Fire-Water (the older system that swaps Rubeus for Fire and Laetitia for Air), but this would be odd considering their pure elemental representations.
• Can other cycles be used instead of the ḥarf cycle?  I know that at least some geomancers use the ABDḤ cycle using the same method of allocating letters to figures, just in a different order of the figures, though it seems the ḥarf cycle is more popular, at least in Africa and the Near East.
• Where did Cattan get his Second and Third Rules of assigning the letters to the figures come from?  I haven’t been able to figure out a pattern there, either, especially with the varied and numerous associations he gives that don’t match anything else.  He even includes the letter Z in the Third Rule!

• Did the methods of determining names as given by Cattan, Heydon, etc. also originally come from Arabic geomancy, or were they developed purely in a Western setting?  If they came from Arabic geomancy, did they come in at an early date and get passed down (and potentially corrupted) as time went by, or were they reimported at a later date?  Given their wording, it seems they were unclear and obscure even in Renaissance times.
• What even are the methods in use for Arabic geomancy for using the ḥarf cycle?  I haven’t been able to read or research much about that, either.  How do Arabic geomancers determine names, and how similar are these methods using the ḥarf cycle (or other cycles with letters associated to the figures!) to those in Western geomancy?
• What can be done about non-alphabetic or non-abjadic scripts?  Syllabaries can feasibly be assigned, syllable by syllable, to the geomantic figures, though that would quickly get out of hand depending on the number of syllables a language has.  How about abugidas, like any of the Brahmic-derived scripts?  How would vowels be handled in that system, if at all?  What about logographic scripts?

Still, even with these unanswered questions, I feel like I have enough at this point to convince me that that whole section in my book’s postscript about how trash these methods of determining names and letters are probably deserves a rewrite.  In fact, what’s astounding about the Western methods is that we have a fossil of Arabic dawā`ir embedded in our own practice, when otherwise there we don’t use any dā`ira-based technique.  It really emphasizes to me that, truly, geomancy is still an art that reaches deep into the sands of north African and Arabic culture, and perhaps there are more things that we can learn from or even merge with from our eastern siblings in this art.

In the meantime, I’m going to get back to more research and writing.  I want to take another look at those rules and try applying them again; now that I have a better understanding of why the letters get allocated to some figures in certain patterns, maybe using the ḥarf cycle in a more pure way than what Cattan or Heydon have could improve those chances of determining names.