Avicenna against the alchemists: the "Aristotle" paragraph

After Arabic alchemy and its Emerald Tablet, this article stays in the Arabic-Persian world with someone who said the opposite of what is usually told: Ibn Sina, Avicenna. In the Shifāʾ there is a paragraph according to which alchemists cannot change one species of metal into another. According to the editors of 1927, it is the passage that begins "Sciant artifices alchimiae" and was quoted "innumerable times" in medieval alchemical literature.² I read it in three languages: the Arabic printed in the critical edition of E. J. Holmyard and D. C. Mandeville (1927), the medieval Latin of that same edition, and the editors' English. I compared them sentence by sentence, and they do not say the same thing. What I did not read, I say at the end.

Who Ibn Sina was

According to the English Wikipedia, Ibn Sina lived from about 980 to 22 June 1037. Ibn Sina was born in Afshana, near Bukhara, in a Persian family, and the works are almost all in Arabic, with some key ones in Persian. Of some 450 works believed to be by Ibn Sina, about 240 survive.¹ The most famous in medicine is the Canon; the one that matters here is the Shifāʾ ("The Healing"), a philosophical and scientific encyclopedia which, Wikipedia says, is meant to "cure" ignorance of the soul.¹

The section on minerals belongs to book V of the Physics of the Shifāʾ. The dates do not quite agree: Wikipedia has the work begun in 1014 and finished around 1020, while Holmyard and Mandeville, citing a medieval biographer, place the chapters on physics after the death of the prince Shams al-Dawla (1021) and before Ibn Sina left for Isfahan (about 1023).¹²

The paragraph, in four versions

The text comes in the edition of Holmyard and Mandeville, which has the Arabic, the medieval Latin and an English translation. The editors say the Latin was translated from the Arabic by Alfred of Sareshel around the year 1200 and added, in three chapters, to the fourth book of Aristotle's Meteorology.² Reading the printed pages, I transcribed the Arabic, the Latin and the English, and translated the Arabic. The piece puts them side by side, sentence by sentence.

What changes from one version to another

The degree. The Arabic does not say "impossible". The key sentence is: "but that the specific difference be removed or imposed, its possibility has not become clear to me; rather its admissibility seems remote to me". The word is baʿīd, "remote". The editors' English renders it "I regard it as impossible"; the Latin, "I do not believe it possible". This is a reader's reading: I am not an Arabist.

The perfect deception. The Arabic says: "I do not exclude that precision may reach a point where the matter goes unnoticed by the sharpest". The Latin does not have that sentence. Ibn Sina admits that the imitation can be so good that it fools an expert, and still says it is an imitation.

The argument from knowledge. This is the deepest part. The qualities perceived by the senses, Ibn Sina says, are probably not the differences that make a metal one species or another, but accidents and consequences; those differences are unknown, "and if something is unknown, how can one set out to produce it or to remove it?" One cannot make or destroy what one does not know what it is. The Latin turns it into another question: how one would know whether it has been removed or not.

The door left open. Right after, the Arabic says that removing or putting on the dyes, the smells and the weights "is something we should not insist on denying merely because we do not know it: there is no proof at all that it is impossible". Ibn Sina does not deny alchemy wholesale. The claim is that changing the species cannot be done as far as anyone knows; changing accidents is not forbidden. In the editors' apparatus there is a Latin manuscript that says here "it is not possible" and in which someone added "im-" between the lines: "it is not impossible".

What the Latin adds and what it drops. The Latin adds "unless perhaps they are reduced to prime matter", which according to the editors the Arabic does not say. And it drops the last sentence: "I have much to say on this, and would say it if I wished; but its benefit is little and there is no need for it here". The editors think the Latin left it out because its sarcasm was "unpalatable"; that is their hypothesis.²

From what mineralogy it is said

Ibn Sina does not reject the theory the alchemists worked with: the one that makes metals out of mercury and sulphur. In the same text, pure mercury solidified by a pure white sulphur gives silver; with a better, whiter sulphur, gold; with good mercury and an impure, combustible sulphur, copper; with corrupt, earthy mercury and an impure sulphur, iron. Of tin the text says that its mercury is good and its sulphur corrupt, hence its "cry"; of lead, that it comes from an impure, heavy, clayey mercury and an impure, fetid, feeble sulphur.²

It is the same theory, with similar rules, that we saw in the Marifetname, and according to Wikipedia Ibn Sina took it from Aristotle and Jabir.¹ The conclusion is this: alchemists achieve solidifications whose qualities resemble the natural ones but are not identical (the editors warn that this sentence is hard to translate), and their art "falls short of nature" and, "in spite of great effort, cannot overtake her".²

Whose paragraph it was

Here the story turns curious. According to the editors, the Latin circulated in manuscripts added to Aristotle's Meteorology, and so many read it as Aristotle's. Albertus Magnus knew it came from Avicenna. Roger Bacon, about 1266, asks that "the fools" fall silent who abuse that authority by saying that "Sciant artifices…" is a word of Aristotle, when everything from "Terra pura lapis non fit" on is an addition by Alfred. Around 1330, Bonus of Ferrara writes that Aristotle "expressly denies alchemy", though "some say those words were Avicenna's, who added them, which we do not believe". And the manuscripts that Holmyard and Mandeville edit close in two ways: "Here ends the book of minerals of Aristotle" or "Here ends the chapter of minerals of Avicenna".²³

In the Muslim world, according to the editors, these ideas are attributed to Ibn Sina unanimously. Ibn Khaldun says Ibn Sina held that metals are distinct species and so cannot be changed into one another, whereas al-Farabi, who took them to be a single species, thought alchemy possible. Al-Jildaki and Hajji Khalifa attribute the denial to Ibn Sina in the Shifāʾ, and a vizier, al-Tughra'i, wrote a whole book to refute those opinions.³

And there is a paradox. In the Middle Ages several alchemical texts circulated in Latin under Avicenna's name. Wikipedia lists four. Anawati, following Ruska, says the De anima in arte alchemiae, the most influential, is a forgery by a Spanish author, and that the Declaratio is not Ibn Sina's either; the Epistola de re recta is less sceptical, and Anawati thinks it is Ibn Sina's, but from before the question was settled in Ibn Sina's mind.¹ The editors of 1927 explain why the authorship was disputed: medieval readers would have found it incredible that someone who denied transmutation in the Mineralia had written other books that accept alchemy and describe the philosophers' stone "with circumstantial detail".² They add a hypothesis: that Ibn Sina made up his mind late, and earlier may have written in favour.

A bird from before

Ibn Sina also wrote a Treatise of the Bird (Risālat al-ṭayr), in symbolic language. According to the Persian Wikipedia, Suhrawardi wrote a treatise of the bird in Persian as a continuation of Ibn Sina's, and used the language of the birds that Ibn Sina, Attar and Ahmad Ghazali had already used.⁴ It is, then, an ancestor of the bird poem of Attar. I did not read the Arabic original, but I did read Suhrawardi's Persian translation, told in the article on Suhrawardi.

What I could not read

I read the paragraph on the printed pages (Arabic, Latin, English), but not in a manuscript, and I am not an Arabist: my reading of the Arabic is a reader's, with one doubtful word in sentence 8, and the rest of the section on minerals I read only in English. I did not read the whole Shifāʾ or the modern Arabic edition. What I tell of Albertus Magnus, Roger Bacon, Bonus of Ferrara, Ibn Khaldun, al-Jildaki, Hajji Khalifa and al-Tughra'i is second-hand, through the introduction of Holmyard and Mandeville, and I read none of them. The Treatise of the Bird I read only in Suhrawardi's Persian version, and the alchemical texts attributed to Ibn Sina I did not read. And the dates of Ibn Sina come from Wikipedia.

What we do with this

Two sentences from this paragraph are useful to us. One: "if something is unknown, how can one set out to produce it or to remove it?" It is the caution with which the article on measuring the world's soul looks at the attempt to measure something nobody can define. The other: one should not insist on denying something merely because one does not know it. Anima Mundis does both at once: it does not measure the soul of the world and does not say it does not exist. What the Observatory counts is the voices that arrived and at what UTC hour, and nothing more.