Reality is a hard truth: “The weight of primary noon . . . the hammer . . . the hard sound . . . the sharp flash.” The word for this reality in Arabic is حقيقة (ḥaqīqah). But poems are full of metaphors and their language is a wind that moves “like a cripple among the leaves / And repeats words without meaning.” The Arabic and Islamic archive is full of poems. It is an archive made up of thousands of memories, books and manuscripts referencing one another across continents, languages, and centuries (elsewhere in this issue, Anna Ziajka Stanton calls it “the premodern compendium of Arabic thought”). Arabic and Islamic poets and critics have had centuries to work through the tensions and ironies that drive the poem I am quoting here, “The Motive for Metaphor,” by Wallace Stevens. Arabic lyric poetry laden with imaginary and imaginative images is the prestige genre that far outweighs all other forms in this archive, but the majority of criticism over its last millennium has skewed toward cold reason and logic. In this short essay, I look for the reality that poets, critics, and philosophers found through poems, and for the truth they claimed lay outside poetry. Stevens, who in a foreign microcosm of that vast dialectic wrote a poem full of metaphoric imagery to criticize metaphor, provides a good way to activate the tensions in English. He wrote about “The obscure moon lighting an obscure world / Of things that would never be quite expressed,” and where Stevens's poem grasps for a truth that he feared could not be expressed, Amal Dunqul wrote a demand reeking of confidence:

The voice of the orphaned daughter in Dunqul's poem claims ḥaqīqah, but when she denies in the very next line that resurrection is impossible (ُولا أطلبُ المستحيلَ ولكنَّهُ العدلُ ُ[“I am not asking for the impossible but for justice”]), the poet and his audience know better. Truth is hard.

But where does it come from? How does one think in literature about truth? about ḥaqīqah? In this 1981 poem, Dunqul was playing with archaisms in direct conversation with a millennium of Arabic lyric culture. Translation requires some brief initial ground clearing: the English word truth combines the difference between the way we talk about speaking truth (as opposed to lying) and the way we talk about the actual truth of reality, the truth that gets reality right. The Arabic and Islamic archive has always used this distinction: الصدق (ṣidq) is the truth of a speech act (the opposite of lies). Ḥaqīqah is truth as accurate connections between language, mind, and reality; between words and ideas; and between ideas and objects. It is also truth as reality itself, what Stevens called “The A B C of being. . . . The vital, arrogant, fatal, dominant X.”

Stevens's X is what I am interested in here; it is the proving ground for literary applications of reason and logic. The stakes are high when truth itself is on the table; ḥaqīqah is both literal (as opposed to metaphoric) language and it is also the metaphysical and cosmological truth of existence itself: وقيل الدنيا باطل والآخرة حقيقة (“the world is invalid, but the afterlife is ḥaqīqah”; al-Rāghib al-Iṣfahānī, Mufradāt 247). Stevens had to resort to a series of images because the simple English word “truth” was not powerful enough for him. In Arabic, ḥaqīqah successfully conjures “the weight of primary noon.”

The truth-versus-metaphor dialectic is a familiar one: Plato complained about the lies of poets; a millennium later the Qurʾan complained about the same problem; Aḥmad Ibn Taymiyyah worried about lyric's ability to lead people astray; and when on the Eastern Seaboard of the United States Denis Donoghue discussed Stevens's poem, he circled back around the Mediterranean to Aristotle by way of Picasso and Nietzsche (548). In this post-Hellenic world (Donoghue and Ibn Taymiyyah both read the ancient Greeks), irrespective of the confines of chronological directions, historical connections, and postcolonial disregard, the hard truth of X stands outside language: “Steel against intimation” (Stevens). My subject matter is the steel, and accounts of “the sharp flash,” in the premodern Arabic and Islamic archive of literary text, criticism, hermeneutics, and theory.

Stevens's truth was an anvil where the hammer hits the steel. Abū al-Ḥasan Aḥmad Ibn Fāris's truth was a spear thrust through the body into the shoulder joint, and Abū al-Qāsim al-Ḥusayn al-Rāghib al-Iṣfahānī's truth was the way a heavy gate post turns in its stone socket. Ibn Fāris and al-Rāghib al-Iṣfahānī were dictionary writers, influential and widely disseminated thinkers-through-lexicography in tenth-century Iran. Just like Stevens, they sought images with which to imagine and communicate; it is not, I think, a coincidence that regardless of genre and context all three men used images of hard, impermeable surfaces of stone and metal when they thought of truth. Ibn Fāris's and al-Rāghib al-Iṣfahānī's images were explanations for the Semitic root حقق or חָקַק , from which the word ḥaqīqah is produced. Al-Rāghib al-Iṣfahānī says this:

He uses a technical term (المطابقة [“correspondence”]) from Arabic Aristotelian philosophy and then goes on to explain it with the image of a heavy gate post revolving in its bushel (حُقّ ): the post remains stable and upright because the post fits and corresponds to the bushel. Ibn Fāris also quoted poetry (2: 15–19):

and

These two lines of poetry by different poets showcase different morphological products of the same root: مُحْتَقٍّ (muḥtaqq; “a cut”) and المُحَقَّقةَ (al-muḥaqqaqah; “something firm”) (the first line is from the seventh-century poet Abū Kabīr al-Hudhalī, the second from an anonymous poet). Just as with Stevens, the truth cuts, and the truth fixes. Here are translations:

and

Ḥaqīqah is an instance of this firmness—a moment of epistemological accuracy in which a relationship cuts straight and true; a judgment is made. The oldest attestations of the word are about protection:

Writing in the eighth century, Bashshār ibn Burd has men protecting female sexuality, while the greatest female poet in this patriarchal archive, Tumāḍir bint ʿAmr al-Khansāʾ, uses the same phrase for that part of her seventh-century community that deserved protection, that had the right to protection, that would rely on her brother for protection before his untimely death:

Ḥaqīqah is a sense of fixedness and rightness, a moral space, a sense of rightness and firmness: people deserve to be protected, things deserve their names, ideas deserve to be described by their names (“deserves” has the same root too: يستحقّ ). These old poetic attestations from the sixth, seventh, and eighth centuries, which became the source material with which scholars and critics like Ibn Fāris would use dictionaries to think about meaning, also recognize the importance of human agency in creating ḥaqīqah: someone had to protect it; someone had to keep it fixed. When the stakes were metaphysical rather than communal, that person would be God. In the ninth or tenth century, the Sufi martyr al-Ḥusayn ibn Manṣūr al-Ḥallāj would say about the cosmological origin of God's love:

Generations of mystics would follow him and use ḥaqīqah to talk about transcendence.

Ḥaqīqah generates both awe and fear, but poetry—ironically—seeks to resolve the threat with lyric imagery. Helen Vendler thought that the voice in Stevens's poem hated itself for attempting such a poetic resolution and feared the hard accuracy that it was trying to praise (Wallace Stevens 23–26). The gendered tension we see in Bashshār's concern for female purity is comparable to the intellectual fear that, while the word may conjure a clean hard accuracy, it remains a subjective decision to call something ḥaqīqah. It is tempting to read a certain degree of hysteria in both Bashshār's and Stevens's series of A's, B's, C's, and X's: Where can poets or scholars go to resolve this fear and give themselves some confidence in their certainty? Not everyone had al-Ḥallāj's confidence that they could see exactly how God created truth, and perhaps some panic shows through the texts of the male scholars and poets during those moments when they face the limits of their control over women, or over knowledge. In the Arabic and Islamic archive, the most popular answer to these problems was reason—human reason.

Qudāmah ibn Jaʿfar's tenth-century discussion of whether “the best poetry is the most untruthful” has attracted persistent and productive attention from scholars inside and outside the Arabic and Islamic archive across multiple languages and centuries (see Harb 35–41; Ajami, Alchemy; Heinrichs 56–68). Here I would like to look closely at the critical responses, which started with Qudāmah, to a particular image in two eighth- or ninth-century lines of poetry by Abū Nuwās:

Qudāmah's objection was that the wine bubbles were compared to light-colored hairs in a black beard and then immediately to black night in a white day, causing a contradiction. He said that black and white cannot be allowed to overlap or blur, because instead of being physical bodies they are categories: a thing cannot be black and white at the same time. Qudāmah dismissed the excuse that the bubbles’ motion rather than their color was the point of comparison because he saw no link between bubbles and white hairs other than their color, and furthermore because the poet had actually chosen to use the words black and white themselves (126–28). Around half a century later, al-Rāghib al-Iṣfahānī's recapitulation of Qudāmah's analysis came in a chapter titled في الجمع بين نقيضَين (“Combining Contradictions”; Afānīn 181–86), in which al-Rāghib al-Iṣfahānī confirmed that poetry has no problem with contradicting its own content for effect, that poets can lie and be good poets at the same time, and that ḥaqīqah—accuracy—can be managed and manipulated by poets: contradictions can be resolved by reading them as referring to separate instances or to different times, or by reading one as ḥaqīqah and the other as metaphor. But if a poet intends two contradictory ḥaqīqah statements that apply at the same time and in the same place—ذلك محال (“that is impossible”; 183). In verses in praise of Hārūn al-Rashīd, Abū Nuwās (again!) cannot in the eyes of the critics say that the caliph is both peerless and at the same time a peer of the (peerless) Prophet Muhammad. Theological sensitivities aside, Qudāmah objected to the combination of negation and affirmation, and al-Rāghib al-Iṣfahānī to the combination of two incompatible ḥaqīqahs (Qudāmah 131; al-Rāghib al-Iṣfahānī, Afānīn 183).

Ḥaqīqah here is managed by human reason and governed by logic. Qudāmah's vocabulary for this work comes from the philhellene and Aristotelian logical tradition, while al-Rāghib al-Iṣfahānī leans into ḥaqīqah. The outcome is the same: when the bubbles of poetry tumble and pop, and poems make one worry about the connection to the hard facts of reality, ḥaqīqah provides something to cling to. But where does it come from? Other than a sense that noncontradiction is a good thing, what is it protecting? At the end of his life, Stevens did not have an answer for what “the thing itself” actually was: “Saying and saying, the way things say / On the level of that which is not yet knowledge” (qtd. in Vendler, Ocean 198). Stevens knew that what people say and have said is, on some level, all we have, but he had no framework to connect all the words to either real knowledge or a structure that might help with the unpindownable nature of poems. Throughout the literary history of the Arabic and Islamic archive, ḥaqīqah was this framework. Its source was communal precedent: ḥaqīqah was what people had said. This appears in the lexicographical work of Ibn Fāris, and one finds it everywhere in technical and literary definitions of ḥaqīqah. Fear and hysteria resolve in the community of language users, their past choices, and the lexicographical enterprise that preserved and activated those word-usage choices. The speech community of formal Arabic, an open and dynamic system replete with (and often created by) nonnative speakers across the divisions of the Islamic empires, but also relentlessly disciplining and elitist, knew it was a literary and scholarly community because it knew what it had said. When thinking in poetry, the ultimate point of reference was the history of language itself. Ḥaqīqah had moved from a word for the people protected to a word for the ideas protected—an idea that itself protects people in a world of poetic destabilization.

The primary language-facing definition of ḥaqīqah is كل لفظ يبقى على موضعه (“words used as they were coined by convention in speech”), or even just ما اصطلح الناس على التخاطب به (“that usage upon which people agree in conversation”; al-Sharīf al-Jurjānī 80). Ḥaqīqah—speech certified by precedent—is opposed to المجاز (majāz), which is the label for all non-ḥaqīqah language (see Noy). Here I must briefly note that the English word metaphor includes both the Arabic majāz and the Arabic الاستعارة (istiʿārah): the former all language that goes outside the bounds of ḥaqīqah inclusive of idioms, dead metaphors, neologisms, implied comparisons, and more; and the latter a poetic image structure in which concepts are borrowed from a source and transferred to a target (see El-Khoury's comparative review of English and Arabic metaphor). This is the mature accounting of ḥaqīqah as language's fixed reference point, taken at a representative chronological midpoint in the long millennium of the Arabic and Islamic archive (al-Sharīf al-Jurjānī 79–80). In the phrase قيد الأوابد (“shackle of the prey”), used by the canonical sixth-century poet Imruʾ al-Qays to praise an effective hunting steed (19), the ḥaqīqah referent of “shackle” is an iron chain, the ḥaqīqah referent of “prey” is the prey, the use of “shackle” to refer to a horse instead of a chain is majāz, and the whole phrase is an istiʿārah in which the action of shackling is borrowed from the chain and given to the horse.

Ḥaqīqah started off as a word for the people in the community who deserve protection, developed into a word for the ideas that deserve to be held as true, and then came to anchor language and reference: ḥaqīqah is literal language and the literal is defined by precedent, what people have said. How does one know that the connection between “shackle” and a restrictive iron chain is ḥaqīqah? Because one can consult the records of language usage in the community—oral or written in dictionaries—and feel confident that one is on stable ground: this is how the word has been used in the world. If this focus on use feels reminiscent of the approach to language made famous by Wittgenstein and Austin, that is because it is: pragmatism avant la lettre by some six or seven centuries (Ali). Although one of course wishes to avoid a teleological account of literary theory as something that develops, scholars in the Arabic and Islamic archive had ample centuries in which to both interrogate the epistemological consequences of a thoroughgoing linguistic pragmatism and integrate their philosophical analyses of reference into literary theory. By way of contrast, scholars in European languages outside the Arabic and Islamic archive have only recently started to argue for the integration of Wittgenstein into literary theory (Moi) or the importance of philosophical models of reference for reading literature (Mohanty).

Scholars in the Arabic and Islamic archive were fascinated by the philosophical tension between a stable account of literal reference and the creativity and dynamism inherent to human language. For example, they were interested in the temporal tension of whether a word became ḥaqīqah when it was initially coined, or whether it became ḥaqīqah when convention settled on it (Çelebi 509–10). In the case of polysemy, they debated the extent to which one considers an original act of coinage or, conversely, subsequent convention when deciding between multiple options for the reference of a particular word (al-Siyālkūtī 479). Ḥasan Çelebi Muḥammad Shāh Ibn al-Fenārī (d. 1481) and ʿAbd al-Ḥakīm al-Siyālkūtī (d. 1657) were working in the Ottoman and the Mughal Empires, respectively, and just as for Ibn Fāris and al-Rāghib al-Iṣfahānī, Arabic was not their first language but rather their language of scholarship and theory: the way they accessed ḥaqīqah. Their mutually referential discussions continued a Mediterranean fascination with origins (the “coining” of words) and development (the “convention” of language), which goes all the way back to Plato's Cratylus in the fourth century BC.

The language science in which Çelebi and al-Siyālkūtī worked filled the shelves of the Arabic and Islamic archive with productive, fine-grained scholastic theories of reference and metaphor (Gleave; Key, “Philosophy”). But that archive also contains a Luther figure, a radical reformer who mastered the complexity and then saw through the fallacies: Ibn Taymiyyah. In the early fourteenth century, Ibn Taymiyyah, a keen reader of the philhellene tradition, its Arabic iterations, and the theological and political history of Islam, decided that ḥaqīqah was entirely made up:

Ibn Taymiyyah recognized that ḥaqīqah was an epistemological category developed by humans, not a reference to timeless ontological truth. He recognized the stakes and decided to rip them out of the ground, completely removing a binary that had structured much of the previous six centuries of thought.

Ibn Taymiyyah left only language spoken in context: speech acts whose meaning was determined, and which could be interpreted at a chronological remove, only through reference to that context and its determination of reference as restricted or unrestricted. If God mentioned جِدَارًا يُرِيدُ أَن يَنقَضَّ (“a wall that wanted to fall down”) in the Qurʾan (18.77), there was no metaphor involved at all. There was just ḥaqīqah, just truth. For in the end even Ibn Taymiyyah could not abandon ḥaqīqah—even he needed a reference point. He could sweep away the distinction between ḥaqīqah and metaphor, and he could locate ḥaqīqah purely in speech acts whose references were determined by their contexts, but he had no raw material other than Stevens's “[s]aying and saying,” and he could not countenance any metaphysical or lexical sense in which ideas deserved their words, let alone the cosmological truth images of a poet like al-Ḥallāj. People are known to have talked about walls wanting to fall down while intending reference to walls that were about to fall down, and this lexical precedent is ḥaqīqah. The whole Qurʾan is ḥaqīqah. There is no metaphor there at all, because there is nothing with which one can make a rational distinction between two species that are figments of the human imagination: the categories of ḥaqīqah and metaphor (Ibn Taymiyyah 7: 107–08).

Ibn Taymiyyah's hard epistemological minimalism brings me right back to where I started with Stevens's anvil of truth and the unease with metaphor. This is not a good place to be for poets, or for the critics who read them. Ibn Taymiyyah has, at least, helped remove some of the fear and hysteria about the sources of ḥaqīqah: one can now be confident that it is just language. Language use determines language use. But simply saying that what appear to be metaphors are actually, rationally, just ḥaqīqah does not help one read poetry. This is where ʿAbd al-Qāhir al-Jurjānī comes in. His eleventh-century theories about metaphor and syntax fundamentally reshaped Arabic and Islamic thought about language. Çelebi and al-Siyālkūtī worked in his tradition, and Ibn Taymiyyah was—at the least—familiar with his work as a grammarian (quoting him twice on questions of morphology [15: 257, 259]).

Al-Jurjānī's ḥaqīqah was the self-sufficient word, the word that when spoken يؤدي جميعَ شرائطه (“fulfills all its own requirements”; 325). This is the same as Ibn Taymiyyah's ḥaqīqah, اللفظ الذي يدل بإطلاقه بلا قرينة (“the word that needs no additional evidence”; 7: 116). But al-Jurjānī recognized that poetry was not a matter of individual words. Poetry puts words together in combination. The fact that individual words stay ḥaqīqah in the imaginary leaps of a poem is what gives the images their power. The ḥaqīqah understanding of what a moon is allows one to make sense of—and feel—the metaphor of an “obscure moon.” When Stevens compares truth to an anvil at work, he invokes a bundle of its characteristics: “the hammer . . . the hard sound . . . the sharp flash.” That bundle is what al-Jurjānī saw as ḥaqīqah—the literal reference of a word is to a group of ideas that constitute the shape and genus of the meaning (Key, Language 224). This thickening of reference to include sets of characteristics was missing in Ibn Taymiyyah's reading of metaphor. Al-Jurjānī's ḥaqīqah could move and fly into the imagined sky of a wholly unreal image—up away from the anvil into the clouds (225). It could fly because al-Jurjānī saw that stars in the mind of a poet are still stars: the bundle of concepts we attach to the word stars remains in play, and that is how the image—and how language—works. I would like to end with two lines of poetry about stars by Imruʾ al-Qays and, from the ninth century, Abū Tammām (for the texts and alternative translations, see Ajami, Pouring 48, 109) and to imagine al-Jurjānī asking us, when we read and see each image, “Is our idea of this star ḥaqīqah?” How real are the stars?