powers, al-Ḥarīrī seems to be grappling with the problem of divine and human language. When God conveyed his final revelation to humankind, he did so in Arabic. With the end of revelation, Arabic becomes a merely human language once again. As such, it can be used to inform, guide, or illuminate, but it can also be used to lie, cheat, defraud, swindle, and deceive. Yet even when it is being used dishonestly, it retains its numinous character: that is, its memory of having once been the voice of the Eternal. 23 Like Milton’s Satan, it retains, even after its fall, some of its original God-given beauty:
. . . his form had not yet lost
All her Original brightness, nor appear’d
Less than Arch Angel ruind, and th’ excess
Of Glory obscur’d . . . 24
It is this numinous character of Arabic that makes verbal miracles possible. It allows Abū Zayd to compose sermons without dots, or verses full of ẓāʾ-words, or speeches that can be read both backward and forward. These are not idle tricks: as Katia Zakharia has argued, games played with a sacred language are never just games. 25 Rather, Abū Zayd’s performances convey what Stephen Greenblatt has called “a pervasive sense . . . that there is something uncanny about language, something that is not quite human.” 26
If we take this tack, a number of things make sense. The narrator, al-Ḥārith ibn Hammām, begins many of the routines by telling us that he went to one town or another in search of some inspiring oratory. This quest appears insufficiently motivated unless we read it as a thwarted reflex of a spiritual search. In late antique Egypt, Christians would journey into the desert in search of holy men, and when they found them, would say, “Give me a word,” meaning a memorable summation of some spiritual precept. 27 This is the sort of word al-Ḥārith is looking for, even if he calls it adab (an Arabic word meaning “disciplined self-presentation” as well as “literary and linguistic training”). 28 Naturally enough, he is drawn to the shabby, hermit-like figure he sees haranguing crowds all over the world. And indeed, Abū Zayd is always up to the task of saying whatever needs to be said as eloquently as possible. Otherwise, there is nothing definite or stable about him: he varies so much in appearance and demeanor that al-Ḥārith almost always fails to recognize him. Abū Zayd may be what Abdelfattah Kilito says he is: a sorcerer’s apprentice who loses control of the forces he has set in motion. 29 But the most economical explanation for his vaporous indeterminacy is that he is Arabic itself. To paraphrase Sheldon Pollock’s description of Sanskrit, he is the language of God in the world of men. 30 And that language is so powerfully in excess of material reality that it overwhelms the agreed-upon relationship of word and object. This unmooring of meaning creates what Daniel Beaumont, one of al-Ḥarīrī’s most perceptive readers, calls the work’s “dreamy, haunting mood.” 31 It also makes Abū Zayd’s manifestations of piety seem forced and unconvincing. By this I do not mean that the real Abū Zayd is a sinner or a hypocrite. As Beaumont reminds us, there is no real Abū Zayd, only “the materialization of a function.” 32 Rather, I mean that when language becomes unmoored from reality, it becomes unmoored from the sacred as well. Al-Ḥarīrī’s language is saturated with the Qurʾan, the Hadith (reports of the Prophet’s words and actions), the rhythms of ritual, and the vocabulary of the religious sciences. But that language is left to fend for itself in a world that seems largely hostile to its purposes, where “the truth is incessantly discovered to be a pack of lies.” 33 Of course, Abū Zayd prays to God to deliver him from poverty and exile, and see him safely to Sarūj, his lost hometown. But the entity that actually defines his life is chance, which is usually malign.
The result of Abū Zayd’s predicament is a desperate search for a passage through or around language. At least, this is one way to make sense of his trajectory. For his part, al-Ḥārith so craves spiritual experience that he is willing to scour the earth “from Ghana to Fergana” (§9.1) in search of words to help him find it. Strangely, though, none of Abū Zayd’s sermons move him to tears of penitence. What al-Ḥārith fails to understand is that the word is not God. His teacher’s speeches are about themselves; the divine must be approached by other means, if it is approachable at all. This is why, despite their humor and occasional raunchiness, the Impostures are suffused with a desperate sadness. 34 Language remains marvelous, but even as we marvel, we know we are seeing an imposture.
Introducing her successful translation of Homer’s Odyssey, Emily Wilson explains her choice of a low-key idiom:
Impressive displays of rhetoric and linguistic force are a good way to seem important and invite a particular kind of admiration, but they tend to silence dissent and discourage deeper modes of engagement. A consistently elevated style can make it harder for readers to keep track of what is at stake in the story. My translation is, I hope, recognizable as an epic poem, but it is one that avoids trumpeting its own status with bright, noisy linguistic fireworks, in order to invite a more thoughtful consideration of what the narrative means, and the ways it matters. 35
Without articulating the principle as clearly as Wilson has, I have always tried to translate in the self-effacing way she describes. But what is a translator to do with an original text whose avowed purpose is to fire off “bright, noisy linguistic fireworks”? Al-Ḥarīrī’s Impostures do not simply include some excessive verbal performances; excessive verbal performance is what they are about. It seems to me, therefore, that any translation that fails to reproduce this feature sells the original short. The problem, of course, is that so many of the fireworks are tied to particular features of Arabic. These include rhyme, especially prose rhyme, and constrained writing—lipograms, palindromes, and the like. Strictly speaking, none of these features can be translated; they can only be imitated. And the only way to imitate them is to throw out the rule book. To understand why, it will be helpful to look at how other translators have treated the Impostures.
In the pre-modern period, responses to al-Ḥarīrī included everything from annotations to word-for-word renditions to imitations to translations proper. The Persian-language responses, documented by ʿAlī Ravāqī, include four translations produced or copied between ad 1191 and 1809. 36 All of them are interlinear, meaning that Persian equivalents for the Arabic words are written between the lines of al-Ḥarīrī’s text. At least one of these translations can be read
