Introduction.

The Work.

The third, and perhaps the most convincing proof that the book is by Nasir, is found in the language and diction of the treatise, together with the terminology. Its diction closely resembles that of the Gusha'ish wa Raha'sh and the Khanu'l-Ikhwan, belonging perhaps to the same period as those books, in Nasir's activities. Nor does its language differ from that of the Zadu'l-musafirin, and, to a smaller extent, of the Wajh-i din. In the case of the latter two such difference may be explained by the fact that in these two works a considerable proportion of the text appears to be either a literal or a very close translation from the Arabic, and this, as usual, very much affects its phraseology and diction. Generally speaking, Nasir's language is very individual, so that it is difficult to think that anyone else would chance to write in exactly his style. To check my own impressions, I have consulted some of my learned Persian friends who are in a position to give an authoritative opinion. All of them, including Dr. Hasan Taqi-Zada, who has done so much work on Nasir, agree that the language is his.

It should be noted, however, that we cannot be certain about the peculiarities of Nasir's language until we find really old copies of his prose works. Those which we possess at present, both those which have been preserved in the Ismaili community on the Upper Oxus and those in the Constantinople libraries, have passed through repeated re-copying by inferior scribes. This particularly applies to the pious Badakhshanis. The Persian of their religious literature is a foreign language to them, speaking, as they do, various local dialects. Being mostly people of very little education, they often misunderstand the text and commit many errors while re- copying it. And, what is much worse, they rarely hesitate to introduce their "correction", sometimes of the wildest kind, which finally upset the reliability of the text.

An interesting feature are the occasional instances of the unusual position of the negation in the sentences, if this is not due to the sentence being a literal translation from an Arabic original. We see on p. 6: ...na az nist ba-hast awurd, or, ibid.: an-chi mar-u-ra dadd bashad na khuday bashad, instead of, khuday na-bashad.

There are many instances of the perfective verbal prefix bi- being used with forms which in modern Persian works do not take it, as in (p. 52) tu bi-rasananda-i (a translation of the Arabic anta mundhir).

There are many cases in which the adjective precedes the substantive contrary to the rules of Persian grammar, or principles receiving the form of the Plural (as, p. 5, afarida-ha). There are many cases in which the form of the comparative degree in the adjectives, with the suffix -tar, is used in the sense of the superlative degree, not implying any comparison.

An interesting word appears thrice (pp. 22 and 35), j-a-k-w-l (jakil, chakil, jagil, chagil, jakwal, jagwal, chakwal, chagwal?), glossed "sharaf wa bala-tar," p.22. I have not found it in any dictionary, and my inquiries in Teheran fromlearned Persians elicited no reliable information.

The other copy (B), undated, is obviously older. It systematically follows the archaic way of writing ch as j, although p is invariably differentiated from b. otherwise there is not much difference between it and (A). The volume, which also includes an unfinished copy of the poem Rawshana'i-nama, consists of 84 pages, 16 by 11 cm., 12 to 14 lines to a page, about 7 to 8 cm long, and is written on a thick brownish Indian paper. It written by several different hands, mostly quite unformed and childish, and has many marginal notes, not connected with the test.

Real variants between the text in these two copies are very rare, although there are very many discrepancies obviously attributable to negligent copying: single words or parts of sentences being omitted, or repeated, or transposed. As usual, Arabic quotations are often so mis-spelt as to become almost unrecognizable, and the orthography of Arabic words is mostly "phonetic".

With all this, although the Shish-fasl produces an impression of exceptional clarity and simplicity as compared with other prose works of Nasir-i Khusraw, the student may find it much below the standard the author wished to achieve. The style is very uneven. In some places the author is so abrupt that it is difficult to understand what he really means. In other places he indulges in endless involved sentences, so interrupted by quotations and translations of the verses of the Coran that he forgets to finish them. Of course, much of this slipshod and negligent style may be attributed to the imperfections of copyists. So may lengthy sentences begin with chunan-ki, or other similar expressions, and it is more than probable that many of these expressions were automatically inserted by scribes who did not properly understand the argument. It is, however, undeniable that involves sentences are a typical feature of Nasir's writings, and form of a great obstacle to the easy comprehension of his thought.

The same applies to his good intention of making his work intelligible to the uneducated by translating Arabic terms. In this respect he is even worse than in the matter of style. Nearly all the really technical theological terms have been left without a translation. This, of course, would be quite legitimate because many of them in Nasir's times had already been well acclimatized in Persian, but amongst those which he has not translated, or even explained, we find some which would hardly be familiar to a non-specialist, as in the case of athar, hadd, etc., while, on the other hand, not much advantage seems to be derived from translating such words as zaman (ruzgar) or such common expressions as khilqat, harakat, etc., by afarinish, jumbish, and so forth.

Thus his good intentions merely create more difficulties because it is usually quite uncertain whether there is any difference in meaning between the Arabic and its Persian equivalent, or whether the author uses this or that word as a technical term or in its plain, colloquial sense. In a philosophical work this, of course, is bound to produce much inconvenience to the student.

In addition to these shortcomings of a purely individual nature, the text, in common with all works on haqaiq, presents much difficulty for understanding and interpretation because of the manner in which the argument suddenly leaps from philosophical matters into the filed of theology, only to jump again into the sphere of mythology, Kabbalistic and other superstitious speculations, and so forth. This, however, is directly connected with the structure of the Fatimid Ismaili doctrine. It would be useful to consider it here, and explain its nature.

Highly developed versions of Islamic theology which make ujse of philosophical speculations for strengthening their appeal to the more educated circles of society, as in Ismailism, and advanced orthodox doctrine, are highly synthetic compounds in which the stratification of the ingredients may often be traced with ease. In all such attempts at analysis we must start with the axiom that all such doctrines were Islam first and last, just plain Islamic doctrine accepted as a whole with all its details, in letter and spirit. All philosophical, superstitious, mythological, mystical and other elements were only auxiliary means to strengthen and support the basic beliefs of Islam. They were used or tolerated, only in so far as they served the purpose. All that could not be reconciled with purely Islamic elements was ruthlessly disowned and rejected, regardless of the fact whether this was consistent with the philosophy adopted by the school, or not. Purely religious elements unreservedly dominated and overruled everything. Thus, when speaking of the "philosophy" of Ismailism, Sufism, and so forth, we must always realize that these were not independent and internally consistent systems, but pieces and fragments of second hand wisdom introduced to amplify this ir that religious dogma or idea. This particularly applies to Ismailism in which such philosophy and references to mythological or Kabbalistic speculations were needed to develop one particular aspect of Islam, namely its theocratic basis in the form of the doctrine of Imamat.

What was Greek philosophy at an earlier period, was by the time Islam began to appreciate it and take interest in it, nothing but an accumulation of debased and popularized knowledge, richly mixed with all kinds of heterogeneous elements coming into Islam almost exclusively through the Christian Church and gnostic sects, it was automatically carrying with it many varied precipitations. Some of these could find a favorable chance to develop, as was the case with the ancient device of allegorical interpretation of myths, scriptures, rites, practices, everything, what in Islamic milieu has come to be known as ta'wil. It was, of course, neither an invention of the Ismailis, nor was it confined only to them, but practiced, often tacitly, even by the most orthodox specialists in Coran exegesis.

The method of ta'wil to a great extent served as a kind of mortar, keeping together pieces of the most heterogeneous extraction and combining them into a finished system. In the Shish-fasl it is used only in very moderate doses, obviously owing to the intention of reducing the size of the work, and especially the author does not deal much with the theory of Imamat, where it is mostly used.

Shish-fasl is obviously compiled from various popular dawat works which were available to the author, but for some special reason he did not plan it on the usual pattern of the haqaiq works of a popular nature, of which we possess many. They form a concise systematic treatise on the doctrine of Ismailism. Nasir wanted to deal here only with the questions of the nature of the soul and its relation with the Godhead, therefore he leaves out many other aspects of the doctrine, especially the theory of Imamat. It is necessary to note that although there are undoubtedly many gnostic and Marcionite ideas implicitly accepted in the argument, the author completely disregards such cardinal problems as the origin of evil and of matter. He does not even mention these important subjects. In his treatment of the theory of creation and the soul, he skillfully takes refuge in purely religious speculations every time he comes to difficult place as far as philosophy is concerned. Certain points remain quite dubious, and raise questions which find no answers.

We see that Satan is never mentioned, the reason for "disobedience" of Soul, their ultimate fate, and so forth, remaining obscure. A great deal is written, with many repetitions, about the relation between the ‘Aql-i Kull and Nafs-i Kull, the part of the human soul in the process of the gradual purification of the Nafs from its primordial defect, etc. The latter, as one may understand, depends on the gradual transference of the elements of the ‘Aql-i Kull to deserving souls in whom alone the synthesis of the elements of the ‘Aql and Nafs can take place. But there is not a word of explanation as to how such independent "influence" (athar) works, obviously over the head of the Nafs? If it is quite automatic in its working, how then comes the discrimination between souls, and why do some of them become "disobedient"?

With this information we may find an easier approach to the rendering of many of these technical terms into English. We may begin with those which appear so many times in the text, ‘Aql-i Kull and Nafs-i Kull, usually translated by "Universal Reason" and "Universal Soul", terms which hardly convey any clear idea. We can easily see that ‘aql, originally "reason", may also imply the object of the reasoning, or the result of it, the "reasoned", or, in application to the expression "‘Aql-i Kull, - "the Rational Harmony of Existence". This would much more nearly convey the idea.

Similarly, nafs, which in Arabic already has so many meanings besides that of "soul", if treated as a masdar, implying the object, may be translated as animation, with (to be noted) strong element of "materialization", this is why the Nafs is the creator of the material world. The tern nafs is also used in many different ways here, as "soul" in the ordinary sense, nafs-i juzwi (individual soul) or as a psychological or biological faculty, etc.

The same may be applied to the difficult expression hadd plur. hudud, which the author uses so much, never giving any explanation. In addition, he uses it in various senses. This term, meaning limit, boundary line, frontier, etc., cannot be rendered by such equivalents, which have no meaning in the context. But it becoes clear if we note that it implies not only the limit, but also the "limited", what it limits, i.e. in our text the "sphere of competence, action, authority." Thus hudud-i ulwi may be aeons, supreme principles of creation, while the hudud-i sufli, in the discriminate use of the author, may mean the phenomena of this world, or also Prophets and other super-men raised by the Nafs in this world.

There is an enigmatic triad jidd, fath, and khayal, which regularly re-appears in early treatises on the haqa'iq, but is never explained. In fact, explanation usually consists, in the works of Nasir and of others, in the statement that are symbolically associated with the angels Mika'il. Jabra'il and Israfil. Despite all my search, I so far have been unable to find any sensible explanation, and it appears to me that here we have to deal with erroneous translation of some Greek or Syriac terms. They undoubtedly of a gnostic origin, and appear in Manichaeism, in exactly the same combination as in here, i.e., belonging to the same order as the ‘Aql and Nafs: Reason, Sense, Intention, Thought, and Imagination. The parallel is complete if we give Nafs the meaning of "sense", which is quite permissible.

It is obviously very difficult to decide in every individual case whether this or that expression is used by the author as a technical term, or not, and whether he attaches some special sense to this or that ordinary word which is quite commonly used. The author, with his haphazard manner of translating certain terms, or leaving them without a translation, does not help us. Sometimes he is so ambiguous and so much corruption has crept into the text that the translator can only depend on intuition. It would be a great achievement to prepare a detailed index of Nasir's terminology from those prose works which have been edited so far. This would help us very much with the edition of further texts.

In order to make it easier for consultation, the terms are here arranged alphabetically.

Afaridgar, khaliq and sani are used to convey the idea of the creator, maker. We may note the "Marcionist" tendency, rigorously carried through, by which God, the One, is never recognized as the direct creator, the maker of the world. This is the task of the Nafs-i Kull. There are, however, expressions such as on p.57: afaridgar ki ‘Aql ast (obviously also indirectly). The verb "to create" is expressed, in addition to afaridan, by ba-hast awurdan, or (52) faraz awurdan. The verb shaktan is not used here in this sense.

Afarinish, not only refers to the act of creation, but also to its result, the creation = the created world. On p.48 an expression occurs afarinish da'if kardan, to weaken the creation, i.e., obviously to obstruct and slow down the tempo of the evolution of the created world.

‘Ahd-i imam-i zaman (p.55) or ‘ahd-i Khudawand-i zamana (p.56), i.e., the oath of allegiance to the Imam, unfortunately for the student, are not commented upon, obviously because the matter does not come under the head of philosophy.

Akhirat, the hereafter, life after death, is not discussed, and here appears only once, p. 34, in the expression ruz-i akhirat, obviously the Judgment Day.

‘Alam, world (only twice replaced by the Persian term jahan, on p. 45: sirat-i jahan, and p.63: in jahan). The term is used very loosely, in a colloquial manner, in the sense of the realm, sphere, region, as (54) ‘alam-i haqiqat, (39) ‘alam-i nafs, etc. there are three ‘alams: the ‘alam-i jismani, or sufli, or kathif, i.e., the material, visible world. ‘Alam-i uluwi, or latif, is the world of the higher realities. And ‘alam-i suwwum (11) which is the ‘alam-i din, the "third world", or the world of religion, obviously simply means the religion of Islam.

‘Amthal, by itself, is very often used instead of the full term, the ‘Aql-i Kull (never Kulli). It is the ‘Aql so to speak, written with a capital A, as against ordinary human reason, ‘aql, which is only on a few occasions specially defined as ‘aql-i juzwi (49), or ‘aql-ha-y-i juzwi (ibid).

Aram (28) is the opposite to junbish, which is Nasir's equivalent for the Arabic term harkat, and thus must be an equivalent of the usual sukun. These two Arabic terms for some reason are avoided in this text. Occasionally aram is also used as an equivalent of rahat, rest, acquiescence, bliss.

Asas, Asasan, is used very often, but never translated into Persian. Incidentally his attributes are mentioned as (52) khudawand-i ta'wil, or (54, 55) nigahban wa parwardgar-i hama hudud.

Asl, asl-ha, is used both in an ordinary sense, and as a term relating to the higher realities, as in (52) asl-ha-y-i ‘alam-i ulwi, obviously the main or the basic principles of the higher world.

Athar, plural athar (aathaar), has been already discussed above. Note the combination on p.17: athar wa fi'l.

Awwal, the First, i.e. the first creation, ‘Aql, Awwal Hasti, or (51) Awwal-i hasti-ha.

Bab is mentioned several times, but only once defined as bab-i Imam (53).

Batin and zahir are here used only on a few occasions, without special explanations.

Bay'at ba-Khuday is here mentioned only once, p.46, and it is not clear what connection it has with the ‘ahd.

Ba-gasht, obviously for the Arabic ruj'at, return, as in baz-gasht-i nafs-i mardum (35), or baz-gashtan-i nafs-i juzwi (36) here only means the final re-union of the individual soul with its source, the Nafs-i Kull, and has nothing to do with any idea of re-birth.

Buzurgwar, great, eminent (26, 27, 35, 44) has here the meaning of Divinely predestined greatness, in the case of great Prophets and Imams, buzurgwar mardum, in whom the Nafs-i Kull attains the peak of its perfection.

Chiz, thing, is used here (14, 58) only colloquially, not in the sense of a philosophical term.

Da'ira-y Imam-i zaman (56) is a great expression which probably does not occur in Arabic works on haqa'iq, and obviously means the period of the office of an Imam, parallel to the dawr of the Natiq. In the Fatimid works it is usually expressed by the tern waqt, or as in old works, qiran. Cf. my "Studies in Early Persian Ismailism", 1948, p.48.

Dar-khurd (7, 13, 36, 45) covers a number of ideas such as relation, harmony, fitting in, coinciding.

Dhat here means not only substance (18, dhat-i ‘Aql-i Kull), but also matter, material substance, as in dhat-i nafs-i ‘Aql (18), where it simply means "flesh".

Gawhar, substance, matter, is also used loosely here: gawhar-kunanda-y-i gawhar (8), i.e. (The one) who has made matter what it is. But there are expressions such as gawhar-i latif (=Nafs) and gawhar-i kathif (='alam), and even gawhar-i nafs-i namiya (8).

Ghayat wa nihayat (51, 53) do not always mean the highest or the finest grade of development of something, but often something into which the object is transformed, as in the sentence: fire is the ghayat of air.

Hadd, plural hudud, has already been discussed above. Here it is often used colloquially, and may be translated by "state", as in hadd-i quwwat, hadd-i fi'l, hadd-i risalat, etc.

Hast alternates with hasti, and it is difficult to see whether this implies any special difference in the meaning, or is due to the negligence or bad Persian of the copyists. Its opposite is nist and nisti. It is not only an abstract notion, as in hast gashtan, but often means creation, beinf as in hast-i nukhust. Hasti-ha simply means being, and something even, as on p. 19, psychial faculties.

Hikmat-ha (48-50) is a colloquialism: wonders, intricate things.

Hisar-i dhat (17), the final limit of properties of a substance, apparently in the sense of the limit beyond which it cannot act.

Juda'i (14-17) here means individuality rather than separate existence.

Junbian, junbish, junbanda, which is so often used, appears to be Nasir's translation of the Arabic verb h-r-k with its derivatives, which he decidedly avoids, only once (27) using the word mutaharrik. We have seen that its opposite is aram.

Kanara, which literally means border, edge, here is strangely used in the sense of designation (58): harf kanara-y-i chiz bashad. Is it an unsuccessful "correction" for kanaya? (kanara), allusion, suggested by the further discussions of the Prophet occupying the position on "the edge" between the two worlds, the lower and the higher?

Kar-kard (4), in the sense of "production", is probably used in accordance with the "Marcionistic" tendency, to avoid expressions such as makhluq, khiqat.

Khudawand is freely used in the sense of the Arabic ‘sahib'. The latter appears only once (68) sahib shinakhtan-i kalima, which is found in the concluding lines of the treatise where traditionally Arabic terms are used for the "pomp".

Kull, the Arabic word meaning whole, entirety, completeness. Here it is repeatedly used in the sense of the source of something, or its genus (jins), as in kull-i lhwish (35, 36, 39, 48, 49). This may be regarded as an individual feature of this text.

Manand is something used as a substantive, instead of manandagi, similarly, not as an adjective.

Maratib-i din (57) is an old haqa'iq term (maratibu'd-din) which does not mean the "degrees" of religion, but its basic principles or dogmas. The word maratib, plural from martaba, here does not appear in its more common sense of a "step, stair, degree", but in the sense of "position, placein, office, dignity," from the original r-t-b, to fix, make firm. It means simply "the foundation of the religion", without any idea of their relative importance among themselves.

Mardum, in the sense of human being, man, is here systematically used instead of ‘adam, insan, etc. Mardumi means humanity (28), and marduman = men.

Markaz-i zamin (54) does not mean the center of the earth, inthe modern sense, but the corn of it.

Miyanaji is here commonly used in the sense of Arabic wasta.

Mu'min here means an Ismaili.

Nafs, like (simply) ‘Aql, is here often used alone instead of the full term, Nafs- i Kulli. Sometimes the latter term is used in the form of Nafs-i Kull, apparently without any change in the meaning. Instead of nafs-i juzwi Nasir uses nafs-i mardum, human soul, and instead of nafs-i natiqa - nafs-i sukhan-guy while nafs-i hissiya and namiya remain untranslated.

Nist, nisti (cf. under hast) in this text very often appears to mean a vacuum, non-existence in a crude material sense, emptiness from material objects. It is this meaning which, in the "Marcionist" tendency of the author, compels him to the theory that God has not created the world from nist.

Nuqsan, defects, is only once replaced by its Persian equivalent (45) ziyan.

Ruzgar on several occasions is used in the sense of "time" 921, 26, 57), although zaman, alternating with zamana, is more common.

Surat is loosely and colloquially used in the sense of either idea or form.

Tamam is often used for tamamat or tamami.

Yaki, with the accent in the last syllable, in the sense of oneness (1, 18), alternates with yakanagi, fardaniyyat.

Zaman and zamana are indiscriminately used in expressions such as Imam-i or Khudawand-i zaman or zamana, apparently without any change in the meaning.