Can anyone decipher the writing around this disc?

Disc with mystery writing around it

It was sent in by a visitor to Omniglot who’s been told that it’s in a Kufic version of the Arabic script, and who thinks it could be Ottoman Turkish or Persian.

A speaker of Hmong who wrote to me recently mentioned that consonants are called tsiaj ntawv txiv (the father letters) and vowels are called tsiaj ntawv niam (the mother letters) in Hmong, and would like to know whether any other languages have similar names for consonants and vowels.

This entry was posted in Language, Puzzles.

28 Responses to Puzzles

  1. Jayarava says:

    In Buddhist esoterism the Sanskrit letter ‘a’ is said to be the mother of all the letters.

  2. Christopher Miller says:

    That piece of information about Hmong is interesting! I’ve been wondering myself if there might be some similar metaphor in the culture surrounding scripts outside Indonesia…

    The indigenous Indonesian scripts are descended, through two different routes, of scripts from South and North India and they inherit the basic structure of their ancestor scripts from India. Their most interesting feature is the fact that there are two kinds of characters: independent letters for consonants and syllable initial vowels, and dependent marks for vowels, which are smaller than the letters and are placed above, below, to the left or to the right of the letters, depending on the mark. The result is that there is a clear distinction between the large autonomous letters and the small bound marks that seem to gravitate around them in various positions.

    In the South Sumatran Rejang-Central Malay script, the letters themselves are called aksara from the Sanskrit word for a syllable, and the marks are called sinjata or senjata ‘arms’/’weapons’. In the Javanese script, the letters are again called aksara but the subordinate marks are called sandhangan ‘clothes’.

    The words used in Batak and South Sulawesi scripts are even closer to the Hmong metaphor. In the Batak scripts, the autonomous letters are called ina ni surat or induk ni surat ‘mother letters’ and the subordinate marks are called anak ni surat ‘child letters’. The same metaphor is used in Sulawesi where the Bugis-Makassarese script is often called lontara’, literally ‘palm leaf’ after the early writing material. The autonomous letters are called ‘mothers of writing’ (ina lontara’ in Bugis, anrong lontara’ in Makassarese), and ‘children of writing’ (ana’ lontara’) in both.

  3. Vivaek says:

    Mother and father letters…interesting….

    In Tamil, we have uyir-ezhuthu (life letters = vowels) and mei-ezhuthu (body letters = consonants).

  4. TJ says:

    About the disc,
    well, to me in fact, seems the design is more in an Andalusian style rather than something Ottoman. The script itself is close somehow to that calligraphy trend that grew in the Arabian west. All in all, it is really hard to read it, but I see repeated letters and curves, and seems to my eye it is only one line repeated all around.

  5. Noam says:

    So interesting! In Hebrew (and Arabic as well), letters all represent consonants, and vowels (which are optional in writing) are represented by a system of smaller marks written underneath or above the letters. Some letters, however, can also denote or stand in for (long) vowels. These letters (aleph, he, waw, and yod) are known in Hebrew as Mothers of Reading: emot qeriah (אמות קריאה), or in Latin matres lectionis. See http://en.wikipedia.org/wiki/Mater_lectionis.
    It is so fascinating to see such a close parallel in a completely different script!

  6. Jerry says:

    Re. consonants and vowels… In Dutch we use the very uninteresting words “klinker” (translates to something like “sounder”) for vowel and medeklinker (“co-sounder”) for consonant. Weird, actually, for a language in which nouns still have a gender (although the correct use of that is slowly disappearing).

  7. d.m.falk says:

    I don’t think Kufic style was ever used by the Turks or the Persians, but was used in Arabic, particularly in what is now Iraq, Arabia and the Gulf states.


  8. Chris Miller says:


    Glad you brought up (mede)klinkers! In the old Dutch descriptions if Indonesian languages of the 1800s, you can see these terms (but ‘meedeklinker’, with double in the old spelling.) Another thing about the Indic scripts is that if there is no specific vowel mark on a consonant letter, then by default it is read with /a/ (or a related vowel depending on the language’s phonology). To show that no vowel should be pronounced, usually at the end of a word/block of letters, but inside a word as well in some scripts, a sign called virāma in Sanskrit is added. In Dutch this is called klinkerdoder ‘vowel killer’. I personally love this word: it sounds slightly funny in English because it’s a short rhyme and because ‘klink’ is so onomatopoetic and ‘dood’ is reminiscent of child-like language: dodo for ‘silly’ and so on.

    But back to medeklinker, which is a direct calque of words based on Latin consonans: the analogy here treats consonants as secondary and vowels as primary, the opposite of the ina/anak terms that are based on the vowel marks looking like smaller letters that gather around the larger consonant letters. (I should point out that the subordinate marks also include signs for certain syllable final consonants including -r and -h as well as nasalisation on vowels — anusvāra or candrabindu –), the Indonesian equivalent being syllable final -ng.

  9. bronz says:

    In Chinese vowels are “original” sounds (元音) and consonants are “complement” sounds (輔音).

  10. dreaminjosh says:

    @ Chris: I was just thinking how silly the word “klinkerdoder” looks to an English speaker. It makes me think “Klinkerdoodle”. Haha.

  11. Max says:

    To: TJ and d.m.falk about the disk

    I am the one who wanted to know more about the language on this disk, so I am very grateful for your ideas. T.J., I went to the sources tonight based on your suspicions that the text might be Andalusian and I was astonished to find that you may well be correct. I found one example of a surviving early 13th century Mahgrebi Andalusian Qur’an page on-line from Jativa, Al Andalus, that shares many things in common with the script on this disc (p.6 of this on line document- http://www.scribd.com/doc/28177243/Arabic-script

    So far this is the closest I have come to a similar script. On the matter of a single text line repeated 6 times, If you look closely, each line is actually different, but shares some characters in common, and is approximately the same length. It is almost as if it were a set of lines from a poetic or poetic/devotional work of six lines. I would love to know what it says!

    I had really hoped that someone connecting with ‘Omiglot’ might possess both the ability to identify the script and also to translate it to English for me.

    On the subject of the artistry I also can find no other similar examples of a work in stone in this form. It does possess a central six sided star and I have read that the very early Andalusian coinage had a star on the obverse side. The number 6 may be significant in that it is repeated throughout the work- the points of the star, the internal divisions, the patterns (2×6 tulips?/ 6 leaves? separating the “tulips”/ and 6 pomegranates?), the six lines of text, what seems to be 6 characters per line/ 6 ‘tulips’ separating the 6 lines of text.

    If the symbols are indeed tulips, it could tie the piece to Turkic (Ottoman) Anatolia, and if the other symbols are pomegranates it may tie the work to the Trans-Oxian region north of Persia, (hereditary homeland of the Turkic Ottomans who later invaded Anatolia).
    I remain very grateful for your insights.


  12. Drabkikker says:

    @ Chris,

    Hey, I didn’t know that one yet. Klinkerdoder, nice. I’m going to try and slip that in a conversation.

  13. TJ says:

    @Max: glad to be of help. Unfortunately, it seems so old to be read easily, but I’m pretty sure the script is Arabic and the style is of the Arabian west. The shapes and letters can denote at some people the word “Allah” and maybe in one position the word “Taha” (another name for Prophet Mohammed). But I can’t really say whether it is a phrase from Quran or so, nor I can tell if it is a common prayer as well.

    As far as I remember, the number 6 has no real significance in the islamic culture. Not that I remember any right now. Maybe 5, 7, 12 or 14 are of more significance. Maybe 4 and 3 as well. These number carry some meanings in some islamic cultures. I think the division of 6 here is purely geometrical and just a design.

  14. Qcumber says:

    Inscription on the disk. Could these be the names of angels?

  15. Max says:

    To: TJ and Qcumber:
    After I read Qcumber’s note, I started to research Islamic angels in a desperate attempt to see if I could find a tie to the text of the disc (I really am grasping at any straws available in attempting to identify the language presented on this disk, and also to translate it to English) and although I came up with no leads specifically concerning angels, one text on Islamic angels did indicate that there are 6 articles of the faith in Islam. The first of these is Faith in Allah, which may account for TJ’s possible reading of one word as “Allah”. The second of these articles of faith is Faith in Allah’s Prophets and Messengers. Could this include the word ‘Taha’ in it, as TJ has read it? I do not speak or read Arabic so I am constantly out on a limb, so (all readers) please excuse any ignorance that I may show about what I read. However I did find a source that listed the 6 articles of Islamic Faith as belief in the following: 1)“Tawhid”(Allah); 2) “Nabi/Rulsul” (Prophets and Messengers of Allah); 3) “Mala’ika” (Angels); 4) “Kutub” (Books sent by Allah); 5) “Qiyama” (Day of Judgment and Resurrection); 6) “Qadar” (Destiny as determined by Allah). The first of these, Tawahid”, is listed as, “Belief in God (Allah), the one and only one worthy of all worship”. Is it possible that both the words that TJ read as ‘Allah’ and ‘Taha’ are in this same line, and may refer to this article of faith? If so, could the 6 lines be listing the Articles of Islamic Faith?

    TJ, I have reviewed every Islamic text that I could get my hands on from the 8th century to the 15th century, from Andalusia, to the Mahgreb, to Timurid Samarquand, and Ghaznavid India, including texts of coinage, architecture, and wood and stone. Of all of them, this text most closely does seem to ressemble early coinage, documents, and remaining stonework from the very early period of Andalusia.

    p.s. I remain at a loss at how to proceed further. She is a silent angel that keeps her own confidence well (smile).

  16. TJ says:

    Qcumber: nope, it’s far from that. Obviously there are short phrases not single words.

  17. Christopher Miller says:

    About the script on the disk: all I can make out is what looks like “lamaa laa lamaa” in Arabic script, repeated in each segment of the circle. I think it’s likely that this is yet another example (of which we’ve seen many on this blog) of Arabic calligraphy executed by a crafts(wo)man who may have been illiterate and not understanding the script, only produced an approximation of the original model. (Among other things, I have in mind the Afghan war rugs we saw earlier this year or last year.)

  18. TJ says:

    @Miller: any idea how to date such a disc?
    I think dating would make somehow a point on whether to consider this a “real” piece to look at, or some arabic-like mimicking.
    I don’t know much maybe about the Ottoman designs, but I’m almost sure that the most “authentic”, let’s say, calligraphy style of that era was the Diwani style. And maybe we know that the most widespread style in the orient starting from Iran and beyond that would be the Nastaliq (or as we call it sometimes the Farsi style).
    The shapes of the letters here I believe can only be seen in the Maghrebi style and the floral designs here are almost a trademark for Andalusian style. Well, Maghrebi and Andalusian were almost one entity back at that time. Hence, we need some sort of dating to make sure that this piece if indeed old and of Arabic origin.

  19. Max says:

    Dear Christopher and TJ:

    Chris I take your point well. I have not been able to pin it down to a comparable piece, although, as TJ has pointed out, its text style does share things in common with the few early Maghrebi and especially early Andalusian documents and coinage that I have been able to compare it to. TJ also raises a good point when he states that if it could be authenticated as to a date, this would certainly narrow down the search. Unfortunately, as with so many ‘historic’ pieces, it is an orphan in time- its past is lost. In fact, if its text can be linked to a specific text style this may well help to locate it for time and geographic origin. This was one of my great hopes, in reaching for help from others far better educated in linguistics than I am. It may be old, or (as Chris observes) it may have been made 30 years ago by an illiterate Afghan banging out the pattern with a rock. This is one of the things I hope to clarify by finding out more about its script and what it may say on it.

    The artistic precision of the work certainly does not compare with the quality and complexity of the surviving early Andalusian works, but this may be simply because most of these are among the rare high quality elite pieces executed for Emirs. The disk does openly seem to show flowers, but this it shares in common with depictions of early Andalusian works, some Turkic Anatolian and Transoxian works, Ghaznavid work, and some north-eastern Islamic Persian work. In fact, it is as difficult to pin down artistically as it is proving to be linguistically. As I stated earlier, she is indeed a lady that keeps her confidences, and her past, to herself. One commonality that seems to be emerging is that both Chris and TJ appear to be able to read different sections of Islamic text from it. This would seem to support a more general agreement that the text does at least contain some Islamic script. My original suspician (based solely on the aesthetic depictions on the disk) was that it most likely was Ghaznavid, or Trans-Oxian, and matched most closely with artisitic styles between the 700 c.e. and 1220 c.e. period for these regions. However it is far more difficult to determine artisitic temporal precision than to make a connection to a script type or translatation of the text. Since I have reviewed the corpus of available surviving material from early Andalusia (as a result of TJ’s observations) I have had to also include this into the possible artistic domain as well.

    One final thought, even if it can be established to be ‘mimicked Arabic text’, this is still valuable to me in that it establishes that at least part of the text is Arabic, and that it was mimicked from an Arabic text. This mimicking may have been done recently, or it also could have been done in antiquity, at a time when Andalusia was in formation, or during the frequent unstable periods of empirical push and pull that occurred as Islam attempted to consolidate its hold on the Trans-Oxus, and on the Afghan territories. Both were times when many people were illiterate to the new language of power- Arabic. If I am able to determine what style of Arabic acted as the ‘master’ for the mimicked text (by the few words that do seem legible), this would also help to establish its parameters for time and geography.


  20. Christopher Miller says:

    Here’s a graphic I made with the six segments of writing from around the disk oriented the same way for easier comparison.


    You can see they all seem to be formed on the same general model, though what that might be is still far from clear. Only the right hand portion of each segment seems fairly constant and at that, there are still differences in each segment. The middle seems to vary the most between segments. And then there’s the curious fact that segment 4 seems to contain text in Latin script: “ovsiqu?”, though that may simply be a coincidental visual artifact that jumps out at someone used to these shapes being used as letters: the ‘S’ and ‘V’ shapes are present in other segments as well without looking as if they are used as letters. Since it’s so difficult to make out any clear Arabic reading, I wonder if this might be some other script entirely? It does look as if it is a right to left script (assuming it *is* a script) just from the way the lines move. Some kind of Syriac? Pahlavi? Something else? (I’m just throwing out wild speculations here…)

  21. TJ says:

    It is getting harder even.
    But as a thought. In #2 there are 3 dots. I don’t remember syriac would use such dots, do they? Hebrew, yeh they use it, but below the letters and obviously this is not Hebrew at all.
    The only script that uses 3 dots on letters and I can think of right now is Arabic. Of course the language could be something else not necessarily Arabic. Unless these dots are for ornamentational purposes, I’d say this letter might be:

    and to make things more complicated, it could be a combination of 2 letters with total dots count of 3. It’s a calligraphic beauty and complication in fact. I’ve seen some calligraphy samples where the dots of 2 letters are mixed up on top, or even below.

    Does the disk has any available history? Is it made of copper?

    Notice: the first symbol in #1 and the last in #5 are the same.
    #5 seems to be starting with Hamza. And there is a big chance that the “S” shape that Christopher was talking about is in fact simply a hamza! The “V” is tricky though, maybe the closest thing I could match it to is “Laa” but also this shape can be seen in Tholoth and Diwani style samples, mainly as an ornament.

  22. Max says:

    Dear Christopher;

    Your idea to use a graphics program to cut, rotate and paste the 6 different ‘lines’ of text, one above the other, was brilliant. I am embarrassed that I did not think of doing this earlier. It allows one to easily compare the 6 ‘lines’ looking for patterns and repetitions of elements. I agree with you. I tend to believe that this text is organized as a flat-bottom-line text instead of a flat-top-line text (such as Hebrew). I also get the same impression that it is written from right to left, rather than from left to right. I will assume that it moves from right to left and use this as a frame of reference as I attempt to identify some repetitions of pattern present in the 6 lines in the following paragraphs.

    A few of the elements (characters) seem to repeat in all, or most of the 6 lines. In the centre of the line you have designated as “1” there is a square U shape, with a straight left side, curved right side “V” shape inside of the squared U. This character occurs in all 6 lines. In lines 2, 5 and 6 it is preceded (right) by an ‘s’ shape. However in line 3 and 4 it is preceded (right) by a different shape. The shape that proceeds the “U” shape in line 4 appears to also proceed the second character in line 1. These seem to have a 3 leaved foliate shape to them.

    The third character in line 3 seems to appear to be repeated in the same position in line 5. It may also repeat in the same position in line 6 but this is difficult to tell for sure due to the quality of the image.

    The character to the left of the “U” shape in lines 1 and 4 appears to be the same except that the shape contained inside of both of these characters is different. In line 4 it is once again the “V” shape mentioned above. In line one it is the “s” shape also mentioned above.

    Line 1 begins (right side) with a reversed “L” shape, while line 5 seems to end with this same shape. The foliate shape mentioned above also begins line 1 and may also end line 5.

    Line 4 and line 6 both begin (right), and end (left), with the “U” shape mentioned above. However the “U” ending line 6 seems to be stretched horizontally in order to contain a series of inner symbols.

    Line 2 ends (left) with an “L” shape unlike any other shape on any other line. It is preceded (right) by a foliate shape, proceded by an “ankh like” shape, and this all proceded by three vertical dots that are also unlike anything else in any other line (except for the terminal foliate shape mentioned).

    Line 1 and line 4 end (left) with a vertical ovoid shape that also occurs inside a central character in line 3.

    I am left wondering if the 3 vertical dots in the centre of line 2 might be for the number 3, and perhaps might be a possible key to cracking the language on this disk (if it is a language). If it is the number 3, it would probably eliminate Islamic as the source of the text, along with many other possible texts. However, it may also help to identify the language being used here.

    It also seems to me that there may be major characters with modifiers attached to them- the smaller shapes enclosed inside of other forms (ie. The “V” in the “U” shape, and the “s”, or foliate shape, to the right of the “U” shape.) Some of these smaller “modifiers seem to occur only once in the text, others accompany different larger characters more than once, and in the same position. Some seem to stand alone, independent of larger characters. Others seem to be intricately connected with larger symbols- some of these enveloping the smaller ones. All of the smaller “modifiers” seem to be restricted to the upper half of the text, while the larger characters seem to all occupy the entire space from top to bottom of the text line.

    Every line has two, and only two, large characters that have a long horizontal stroke parallel to the bottom delineation of the text. All lines seem to possess one, and only one, basal curved larger character. In lines 3, 4, 5 and 6 these basal curved characters seem to be in the same position in the line, but each is individually varied, either by its shape, or by the ‘modifiers’, or by both. Line 1 also possesses a single basal curved shape, but it is at the end (left) of the line. Line 2 is unique in that it only possesses a “J” shape, accompanied by the 3 vertical dots mentioned above.

    All lines appear to have only 3 large characters with a variety of smaller “modifiers’ scattered in and around them. Line 2 may be the only line that has 4 or 5 large characters (if the “J” shape, the 3 dots, and the “ankh shape” are considered as separate characters.)

    I am left wondering if your reflections concerning the origin of any text that may be contained here could be instinctively correct. Perhaps it is not at all Arabic. My instinct tells me that these are not random shapes manufactured to similate text. It seems to me that there are too many internal patterns, and subtle mutations, and seemingly intentional repetitions and variances to indicate a meaningless psuedo-text.

    I am not an expert in antique languages, and therefore I do not have the background to allow me to recognize similarities that would lead me to specific ancient languages, but I have previously compared the text characters to examples of both Syriac and Pahlavi text and characters (and their derivatives) and did not find similarities. The closest I have come to date is actually early Andalusian/Mahgrebi text, although even here only a few of the characters seem to match, and the overall text is quite different in appearance. I am inclined to suspect that your educated speculation that it may be “Something Else” may be correct. My problem remains…….. What else? I remain at a loss for how to solve this problem, who to turn to, and where to look next. Still, the individuals at Omniglot have attempted to help me find a resolution to this problem and have given me far greater insight into the possibilities and the nature of the problems at hand. For this help I am grateful. I had hoped that someone would have written, “Oh…… What you have here is ……..? It dates to this period,………., and comes from this region,……….”. Oh if only life were that simple (smile!). Still, with each new day, and each new insight from others I feel I may be getting closer to a Eureka moment. (Hope lies eternal!)

    Christopher, I am very grateful to you for your observations and reflections, and for your work rotating and matching the 6 lines into one graphic. Surely in this form someone will recognize it, if that is possible.


  23. Max says:

    Dear TJ

    Maya and Nahuatl use three dots, one above the other, to denote the number 3. Although this is definitely not a product of any culture from the Americas, do you know of any culture from the Middle East, Near East, or India that used a dot notation for the number 3? This could provide an important clue to the origin of any language presented here since that it seems to me that most Middle Eastern, Near Eastern, and Indian cultures used abstract character symbology to express numbers. This text may come from a culture and language that used a dot representation for number 3, just as the Maya did. This may help us to identify the language, if this is indeed a representation of the number 3. (And that is only a supposition.)

    You ask, whether the disk has any known history, and, “Is it made of copper?”. The answer to the second question is no. It is made of a clear white crystalline stone that seems to be a quartzite, calcite or marble. The stone itself takes a very fine polished surface even though the carving on it is not highly refined. The surface appears brown due to a heavy patina of exposure to earth, and appears to have been outside for a very long time. The back is rough shaped, flat, and wears the same earthy brown patina. When originally carved it would have had a glistening white surface, similar to a Ghaznavid (c. 1100 c.e.) marble basin I saw recently for auction at Christie’s. It is about 22 cms (9 inches) in diameter and is about 6 cms (2 ½ inches thick).

    To answer your first question, the only history that is know about the piece is that it was reputed to have originally come from the region of Afghanistan, although this may only be a created history from someone that was not sure. To be honest, its past has become detached from itself. That is what I am attempting to do……. find out more about its origins by way of the stylistic elements on it and the language used. The best bet for me is to use the language to tie it down in time and place, however (as you have so correctly observed) this is also turning out to be a far from simple task.

    I hope that this additional information can help in some way.


    p.s. One other thing that occurred to me. If this is not an Islamic type script it most likely predates Islam, or comes from an area that lay outside of Islamic control in-so-much as so much of the Middle East adopted Islamic script as the lingua-franca after 700 c.e.. This might also lead one to suspect that the language represented here is one that was used in the Middle East prior to 700 c.e., or that was geographically peripheral to the Islamic controlled domains after 700 c.e..

  24. TJ says:

    @Max: As far as I remember, I think the Sumerians (or babylonians) used a numerical system with 3 dots as the number “3″. Unfortunately I’m out of the country at the moment and I can’t reach my books so far.
    Although I’m not quite familiar with ancient designing patterns in central Asia specially in eras before the expansion of Islam to that region, yet, I do feel that it is something related to the west and not the eastern parts of the world.

    In #2, which Miller thankfully prepared, the last 2 characters (or the first from the left) can be read as “T^aa”, with “T^” is a hard plosive palatal sound (not sure how to describe this sound). The Shape of this letter specifically, is almost Maghrebi in style. However, no clue about the language.

    Thinking about it even further now, since we have shapes similar to “S” and “V” and even “O”, what about Old Slavonic, Cyrillic (specially the old one) or even Coptic?
    Another option I was thinking of is, what if this disc is supposed to be a seal? i.e. we’ve been looking at the reversed image of what is supposed to be (like how seals are usually made)?

  25. Max says:

    Dear TJ

    WOW! Now that is what I call being able to think abstractly out side of the box. The minute that I read your final comment about the fact that this may be a seal, and therefore we could be attempting to read the text backwards it hit me. I have seen seal stones such as this from the Afghanistan region before. They are also often made of the same type of stone. It would also explain why every surface image on the disk is cut to a single raised plane….. in order to take the ink and transfer it to the new surface in a lithographic process of print making. This would also explain the size of the piece, at a diameter of only 23 cms (9 inches). And the fact that the carving is restricted to simple surfaces and spaces between them with little of the embellishment that normally accompanies stone sculptural work. I am sure that you are correct in this. I have reversed the disk image here so that it will appear as it would if it were lithographed onto a new surface (ie reversed). Does it make any more sense this way???


  26. TJ says:

    This is interesting even more now… although unfortunately I still can’t read it. But who knows, maybe someone with more experience than I do will make sense of it. When I see things in reverse now, the shapes are less likely Arabic ones.
    Seems you’ve worked a lot with Afghan artifacts and hence I believe it’s in your hand now to identify this more, specially that you identified the style and the purpose and the make of the disc itself!
    I’m still puzzled though, more often, by the S V O letters.

  27. Max says:

    Dear TJ, Christopher and all others:

    I just wanted to take this opportunity to thank all who have contributed to this blog strand. I have learned a great deal from you, and have benefited greatly from your experience. Although I still do not know whether or not this is a script, a psuedo-script, or an invented script, its place or origin, or its time of creation, I do know a lot more about what it is not, and what it may be. I have found the discussion concerning this ‘orphan of time’ fascinating. It remains an illusive puzzle to me- one that I hope one day to solve. If anyone does have a moment of clarity, or illumination about the script, I would still appreciate hearing about it


  28. D O'Donovan says:

    Dear Max,
    I hope you have worked out the script already, but here’s one thought. If it is a script from the Maghreb, then depending on the time it was made, it might represent an effort at formalising a Sabaic/Himyarite script. The last rulers of Islamic Spain were from the Azd tribe, who traced their history to that region, and we have numerous examples of medieval “Himyarite” scripts which resemble exactly none of the older ones, but a combination of several. Okasha El-Daly refers* to some manuscripts in which these scripts can be seen.
    *Egyptology: the missing millennium