Wednesday, March 14, 2012

Hahalruna and Persian Khatt-i-Shajari

 I recently wrote about Hahalruna and El-Mushajjar, showing how the Arabic tree writing was a perfect port of Norse hahalruna (cipher runes) into Arabic writing. While writing that article I became aware of a very similar Persian cipher called Khatt-i-Shajari (Tree writing). Just as the Germanic people devised many variations based on the concept behind hahalruna (iisruna, lagoruna, etc), so too did the Persians. The Persian adaptations include smoke signals, hand signs, and tapping that sounds oddly like Morse code. I have read that the Celts used similar methods to communicate their Oghams. I see no reason to believe the Old Norse did not also communicate their runic ciphers in just as many creative forms. Even if they did not, perhaps modern Rune Masters should.

Although Sir Richard Burton and others mention this Persian code, the best description I have found is in A Year Amongst the Persians:
impressions as to the life, character, and thought of the people of Persia, received during twelve month's residence in that country in the years 1887-8. (Edward Granville Browne, published 1893) :

While I am on the subject of these linguistic curiosities, I may as well mention a method of secret communication sometimes employed in Persia, the nature and applications of which were explained to me by my Erivani friend a few days before his departure for Mashhad. Such of my readers as have studied Arabic, Persian, Turkish, or Hindustani will know that besides the ordinary arrangement of the letters of the Arabic alphabet there is another arrangement called the "abjad" (from the four letters alif, ba, ,jim, dal which begin it) representing a much older order. The order of the letters in the abjad is expressed by the following series of meaningless words, consisting of groups of three or four letters each supplied with vowel-points to render them pronounceable :—abjad, hawaz, hoti, kalaman, sa'fas, karashat, thakhadh (sakhaz) dhadhagha (zazagha). In this order each has a numerical value ; alif= 1, ba = 2, jim = 3, dal = 4, and so on up to ya = 10 ; then come the other tens, kaf= 20, lam=30, and so on up to kaf= 100; then the other hundreds up to gheyn=l000. The manner in which, by means of this abjad, words and sentences may be made to express dates is familiar to all students of these languages, and I will therefore only give as a specimen, for the benefit of the general reader, the rather ingenious chronogram for the death of the poet Jami, premising that he was a native of the province of Khurasan ; that "smoke" or "smoke of the heart" is a poetical term for sighs; and that to "come up from" in the case of a number means to be subtracted from.

This, then, is the chronogram: "Dud az Khurasdn bar amad," "Smoke (sighs) arose from Khurasan," or "dud (dal = 4, vav=6, dal = 4 ; total 14) came up (i.e. was subtracted) from Khurasan" (kha= 600, ra= 200, alif= 1, sin =60, alif= 1, nun = 50; total 912). Taking 14 from 912 we get the date of Jami's death, a.h. 898 (= a.d. 1492).

The method of secret communication above alluded to consists in indicating first the word of the abjad in which the letter to be spelt out occurs, then its position in that word. In communicating by raps, a double rap knocks off each word of the abjad, while on reaching the word in which the desired letter occurs its position in that word is indicated by the requisite number of single raps. An instance will make this clearer. It is desired to ask, "Nam-i-tu chist ?" (" What is thy name ?") : the letters which spell out this message are—niun, alif, mim, ta, vav, jim (for chim), ya, sin, ta. Nun is in the fourth word of the abjad, and is the fourth letter in that word (kalaman). It is therefore indicated by three double raps (removing or knocking off the three first words, abjad, hawaz, hoti, and thus bringing us to the next word, kalaman), followed by four single raps (showing that it is the fourth letter in this word). The remaining letters are expressed in similar fashion, so that if we represent double raps by dashes and single raps by dots, the whole message will run as follows: — — — • • • • (nun); (alif) ; — — — • • • (mim); — — — — — • • • • (ta)  • • (vav); • • • (chim or jim); — — • • • (ya); — — — — (sin); — — — — — • • • • (ta).

Messages can be similarly communicated by a person smoking the kalyan or water-pipe to his accomplice or partner, without the knowledge of the uninitiated. In this case a long pull at the pipe is substituted for the double rap, and a short pull for the single rap. Pulling the moustache, or stroking the neck, face, or collar (right side for words, left side for letters), is also resorted to to convert the system from an auditory into a visual one. It is expressed in writing in a similar fashion, each letter being represented by an upright stroke, with ascending branches on the right for the words and on the left for the letters. This writing is called, from the appearance of the letters, khatt-i-sarvi ("cypress-writing") or khatt-i-shajari ("tree-writing"). In this character (written, in the usual way, from right to left) the sentence which we took above ("nam-i-tu chist ?") will stand as follows :—

From the description given by Mr. Browne I was going to make a nice table with Persian letters and their khatt-i-shajari equivalents. The only problem is I did not know what letters to use. Sir Richard Burton gives us a version of the cipher (HERE):
As with the cipher preceding it in Burton's article, this cipher only has 22 letters. Of this Burton writes "it contains only the ancient and universal Semitic letters, lacking the last six of Arabic". Sir Burton then tells us the cipher I just showed you above "is applied to Pehlevi or old Persian". When I did a google search for Pehlevi and Old Persian I got about twenty different scripts and not knowing much about Old Persian, I didn't want to make a guess at which one might be correct. 
From the information supplied by Mr. Browne I see that the Arabic letter jim is used in place of the Persian letter chim. The other variant letters probably also use the Arabic letter they resemble. So, using the information supplied by Mr. Browne, I have retrofitted the old Arabic cipher given by Ibn Wahshiyah simply by removing one limb from the right of every tree in the El-Mushajjar cipher, thus:

As you can see, the Persian version of the cipher is a slightly more economical variation of Arabic tree writing.
Anyway, it seems I have gone another night without sleeping. If you have feedback or more information on any of the codes and cyphers I have written about, please let me know.

Creative Commons License
Hahalruna and Persian Khatt-i-Shajari by Jason Bales is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License.
Based on a work at