Part V

What Was the Abbasid Caliph’s Library Called?

The sources at hand do not have sufficient information to ascertain when the first library in the Abbasid court was established, however at the outset it is safe to say that the beginning of such a library would accompany the translation movement and scholarly activities, which started with the personal attention and interest of al-Mansur (r. 754–775), the second Abbasid Caliph and the founder of the city of Baghdad, and his circle. According to Sa’id al-Andalusi (d. 1070), al-Mansur was the first Abbasid Caliph to cultivate science and in addition to his profound knowledge of logic and jurisprudence, he was very interested in philosophy and observational astronomy and the people who worked in these fields.¹ This keen interest in science and learning no doubt acted as a kind of motive to establish a library in the palace. According to Katip Celebi (d. 1657), al-Mansur asked the Byzantine emperor to send him science books and the emperor sent him Euclid’s books and others.² Al-Masudi (d. 956) tells us that “Once in possession of these books, the public read and studied them avidly”.³ The available information in the narratives of the early Abbasid period is explicit about the existence of the palace library during the reigns of Harun al-Rashid (r. 786–809) and his son al-Ma’mun (r. 813–833). This information is mostly found in the al-Fihrist of al-Nadim (d. 990), as well as in other historical accounts relating to that period. According to these sources, the presence of a library attached to or within the palace complex during the reigns of Harun al-Rashid and al-Ma’mun is a well-established fact, as will be seen in the next chapter.

Al-Nadim, in his frequent references to this library as already explained, uses different names made up of a combination of the words Khizana (repository), Bayt (house), Kutub (books), and Hikma (wisdom). It is obvious that he refers to the same library, which belongs to the same caliph, with more than one compound name composed from these words. For example, in one place he states that Sahl bin Harun was in the service of al-Ma’mun as the director of Repository of Wisdom (Sahib Khizanat al-Hikma),⁴ while in another place he mentions that he was the director of the House of Wisdom (Sahib Bayt al-Hikma).⁵

Al-Nadim, in his short references to Sahl bin Harun, Said bin Huraym and Salm, notes that Sahl entered the service of al-Ma’mun and became the director (sahib) of the Repository of Wisdom “Khizanat al-Hikma”. Later he refers to Said bin Huraym as the partner sharik of Sahl bin Harun in the House of Wisdom “Bayt al-Hikma”.⁶ As for Salm, he says twice that Salm was the director of the House of Wisdom⁷ and once said that he was the director in association Ma’ with Sahl bin Harun.⁸ All these imply that there was more than one administrator at the same time at Bayt al-Hikma. Meanwhile, he clearly mentions al-Ma’mun’s library by two different names.

Although the above-mentioned two terms refer to the same institution, we understand that the compound name Bayt al-Hikma was used mainly for the caliph’s library, distinguishing it from the other libraries or any repository for books. The literal meanings of the words “khizana” and “bayt” as explained above show the difference in identification and capacity.

Al-Nadim in general uses the name Khizana for the libraries of Abbasid Caliphs (e.g., Khizanat al-Mu-tadid) excepting those of Harun al-Rashid and al-Ma’mun.⁹ Surprisingly he takes great care in using the different names for al-Ma’mun’s library. In writing about the scholarly activities in al-Ma’mun’s time, he uses the names Bayt al-Hikma¹⁰ or Khizanat al-Hikma;¹¹ however, when he refers to what remained from this library in his own time, while he is listing the books that belonged to this library, he calls it Khizanat al-Ma’mun.¹²

His “selective appellation” is not an oversight or due to his usage of synonyms. It is evident from the information we have at hand that calling the palace library Bayt al-Hikma in the periods of Harun al-Rashid and al-Ma’mun was not only because they housed books related to what has been called wisdom but also due to their housing the kinds of scholarly activities related to topics in these books of wisdom. For al-Nadim, who wrote his work one and half centuries after al-Ma’mun’s death, what remained of this library was no different from the other libraries that collected books including books on wisdom without being the venue for scholarly activities that distinguished the two caliphs’ library from the others.

It is a well-established fact that the two caliphs, who owned this library known as Bayt al-Hikma, were after the books of the pre-Islamic cultures that were in their lands and in the neighboring countries outside their domain, specifically the books in the Byzantine collections. Considering their close interest in the history and culture of the Greeks and Persians, it is plausible that they were aware of the famous ancient libraries belonging to these two worlds. However, we do not possess information to verify whether they took the ancient libraries as models in establishing their own, as maintained by farfetched assertions that we come across in modern scholarship. There is no evidence that this term was translated from Greek or Persian, furthermore, those who maintain this viewpoint have not even provided the Greek or Persian equivalent of this compound name or even of the word “library” in those languages.

Youssef Eche, who was one of the first scholars to study the history of Bayt al-Hikma in detail, maintains that the Arabs began to build libraries similar to those found in the lands that they conquered and called them Bayt al-Hikma or Khizanat al-Hikma. Without giving any historical example or proof, he claims that this appellation is the Arabic translation of the names of pre-Islamic institutions. Eche also alludes to what he considers as similarities between the Abbasid Bayt al-Hikma and the Museum of Alexandria and draws parallels between the two. He supports his suggestion by anachronistic evidence taken from al-Biruni (d. 1050) and al-Maqrizi (d. 1442), based on their accounts of Greek and ancient Egyptian institutions.¹³ The two quotations that Eche refers to as Bayt al-Hikma and Dar al-Hikma do not substantiate the claim that these were the original appellations of pre-Islamic institutions, as much as they give good examples for the common usage of this compound name as an established name for institutions dealing with sciences related to pre-Islamic civilizations.

In related Arabic literature there are many clear examples that show the term House of Wisdom after it was coined and became frequently used to denote ancient Greek institutions dealing with science and philosophy. The famous Abbasid author and translator Hunayn Ibn Ishaq (d. 873), a century before al-Nadim, in his book Adâb al-Falâsifa writing about the meetings of Greek philosophers, says that they met in one house of the houses of wisdom (Beyt min buyut al-Hikma) to discuss philosophy, letters, and wisdom.¹⁴ It is clear that he is using this newly coined Arabic term as an attribution for the Greek institution where Greek philosophers would meet. In Part VI, we will elaborate on this point further.

The famous geographer Ibn Hawqal (died after 973), who lived in Baghdad, briefly mentions Rome and Athens in his work Suret al-Ard. He calls Athens “Dar Hikmet al-Yunaniyin”; that is, the house or seat of wisdom of the Greeks.¹⁵ In his brief explanation, Ibn Hawqal states that “sciences and philosophy/ulum wa Hikma” of the old Greeks were preserved in Athens, which was considered to be the house of wisdom. It is interesting that Ibn Hawqal uses the word “dar”, a synonym of “bayt” which specifically means a large-sized mansion, a house comprising a court or a set of apartments.¹⁶ Contrary to Ibn Hawqal, al-Nadim in his different accounts on the Greek philosophers and Aristotle and their teaching philosophy at the Lyceum in Athens, uses different appellations and calls the Lyceum the house of education, “Dar al-Ta’leem”.¹⁷ For instance, fourteenth-century historian and geographer Abulfeda (d. 1331) in his well-known book Taqwim al-Buldan refers to Athens as the city of Greek philosophers that is to say the “Dar al-Hikma” of the Greeks.¹⁸

Obviously, tenth-century authors and those in the following periods, other than al-Nadim, were using the generic name of Bayt al-Hikma, which was previously coined for the libraries of Abbasid Caliphs, for pre-Islamic royal libraries. Its synonym Dar al-Hikma, connoting a wider place or meaning a seat of wisdom, was used for Athens, which they considered an important seat of philosophy and the sciences of the ancients.

In an attempt to further elaborate on the history of coinage of the appellation the House of Wisdom, it is rather a rewarding exercise to compare two texts related to the same book written at different times. This comparison provides an important clue to the emergence and development of the appellation of Bayt al-Hikma as an Arabic terminology for the royal library. Ibn al-Muqa’fa (d. around 756), the famous author and translator, who came from Persian descent and pioneer of Arabic literary prose, translated the famous collection of tables from Indian roots known as Kalila and Dimna from Pahlavi into Arabic.¹⁹ In this book, there is an autobiographical introduction written by Buzoe for the translation of the book from Sanskrit to Pahlavi. Buzoe was a physician who was sent to India by the Sassanian King Khusrav Anusharwan (r. 531–579) in search of this famous book. Ibn al-Muqa’fa translated the voyage and adventures of Buzoe in India, which included references to the library of the Indian king. Here, Ibn al-Muqa’fa refers to the royal libraries as “khaza’in” which is the plural of “khizana”.²⁰

The second text is an introduction of extraneous origin and ascribed to a certain Bahnud Ibn Sahwan, also known as Ali Ibn al-Shah al-Farisi, where he gives an account of the history of the book in India. French orientalist De Sacy (d. 1838) published this introduction together with the text of al-Muqa’fa. However, this introduction is not found in the oldest four manuscripts of the Arabic text; most probably it was added later. The date this introduction was written is not yet established but it is clear that it was written long after Ibn al-Muqa’fa’s time and was included in some later manuscripts. Here, the short form of khizana/khazan’in was not used in referring to the library of the Indian kings, as was the case in the early authentic text, which Ibn al-Muqa’fa translated from Pahlavi.²¹

Instead, the appellations Khazain al-Hikma and Bayt al-Hikma are used here.²² We may conclude that during the time the second introduction was written, the terms “Bayt al-Hikma” and “Khazinat al-Hikma” were already developed as compound names for the royal library. They were used after the death of Ibn al-Muqa’fa in the middle of the second century AH/eighth century CE. They were used frequently in this context when the introduction attributed to Ali b. al-Shah al-Farisi was written.

As an antithesis to Youssef Eche’s Greek allusion, Dimitri Gutas in his study on Greek Thought and Arabic Culture assumes that “in the first place, Bayt al-Hikma as a term is the translation of the Sassanian designation of a library”.²³ This assumption is advanced in a context where he tries to relate this institution to what he considers as the attempts of the Iranian bureaucrats of the early Abbasid epoch to restore the Sassanid administrative and bureaucratic setup and recover the Sassanid imperial ideology. The discussion of this assumption falls outside the scope of this study; however, what should be emphasized in this context is that, while Gutas negates the affiliation of Bayt al-Hikma with Greek legacy, he tries to establish its relation to a pre-Islamic Persian past. As implied evidence for such an assumption, he quotes a sentence mentioned in a book on poetic proverbs entitled Kitab al-Amthal al-Sadira ‘an Buyut al-shi’r written by Hamza al-Isfahani (d. after 360/970).²⁴ The manuscript of this book consists of an introduction, two divisions, the first divisions consist of seven babs and these comprise 60 fasls while the second division is composed of seven topics (fen). al-Isfahani elaborates on the Arabic poetic proverbs and states that in this context the Arabs have preceded other nations to which he makes short references. In his reference to Greeks, he says that their liking for poetry was less than that of the Arabs and Persians and refers to the Greek scholars’ poetic quotations in their writings. He mentions Aristotle who liked poetry and quoted Homer. He also refers to Plato who did not like poetry and considered that poets and painters were deceived by their ears and eyes respectfully. Al-Isfahani had a collection of 4700 Arabic distiches and hemistiches that became proverbs among the ancients Arabs and approximately 400 proverbs were translated from Persian to Arabic. In mentioning poems in pre-Islamic Iran, he says the books that contained Persian lore, war reports and various pieces of information about famous lovers that were originally composed in prose, were recast in poetry for the [Sassanid] Kings. In this context, al-Isfahani continues to say that these poems were written in books that were to be “kept forever in the repositories (khaza’in), which were houses of wisdom” (tudawwan fi butun al-kutub wa tukhallad fi al-khaza’in allatî kanat buyut al-hikma). Al-Isfahani concludes by saying, “With the decline of their state, most of these books were lost; however more than ten thousand folios of writings and poems in Persian were left behind”.²⁵

Gutas drew his above-mentioned assumption that Bayt al-Hikma as a term is the translation of Sasanian designation for a library and he notes, not from the unpublished manuscript of al-Isfahani’s book but rather from “the summary information given by Gregor Schoeler”.²⁶ It is obvious that al-Isfahani, while calling the Persian royal libraries “houses of wisdom”, does not explicitly or implicitly refer to its Persian name or to the appellation Bayt al-Hikma as its Arabic translation. He was simply using the established name for a royal library in his time (tenth century).

Meanwhile, Ibn Hamza al-Isfahani, in his published book on the “History of Kings and Prophets”, provides clear evidence about the interchangeable use of the term “Khizana” for the court library of al-Ma’mun. In the first chapter of his book, he refers to eight books on the history of Persian kings that he consulted and mentioned a certain book, which he says was copied from al-Ma’mun’s library; “Khizanat al-Ma’mun”.²⁷ It is clear that he calls this library after the death of its founder in the same manner al-Nadim does by saying Khizanat al-Ma’mun.

Our research could not locate any specific Pahlavi term that denotes the name for a royal library. Meanwhile, on the authority of scholars of Middle Iranian language such a term does not exist in surviving Pahlavi texts.²⁸ However, in support of his above-mentioned assumption in a later published article on the subject, D. Gutas and Kevin Van Bladel mention that the expression of Bayt al-Hikma (as well as the alternate expression, Khizanat al-Hikma) apparently is the Arabic translation of a middle Persian term ganj for “libraries of Sasanian kings”. This assumption is based on a study of M. Shaki namely the Denkard Account from the sixth century of History of the Zoroastrian Scriptures, dated from middle of the sixth century.²⁹

On his elaboration on the word ganj mentioned in this text from the sixth century, and its different readings, Shaki provides a detailed discussion which does not reflect any similarity between the Arabic term of House of Wisdom and the Pahlavi term ganj. He clearly says

… These considerations apart, we do not know of any “satrapal” [the governors of the provinces of ancient Median and Achaemenid Empire] treasury or archives which, none the less, were rendered “royal” by the scholars who have accepted this reading. The oldest reference to such a state or royal archives is made by the Bible … They preserved the Avesta and Zand which were both written in gold on cow hides in the Royal Treasury. Bôxtagân states that he deposited his ayâdgâr in the ganj ı sahigan, which shows that this Royal (Palace) Treasury was also used as archives.³⁰

D. Gutas in his book Greek Thought and Arabic Culture, quotes this text of Denkard’s book and endorses the explanations of Royal Treasury.³¹

It is obvious from this sixth-century Pahlavi text that the word ganj is used to denote the place where books or other archival material would be kept under the Sasanian rulers. However, this has nothing to do with the compound name of the Abbasid House of Wisdom. It is clear that ganj is equivalent to the Arabic Khizane, and at the same time the Greek thêkê, as both denote the place or the case where books are kept. Meanwhile, it is satisfactory to note that ganj in the new Iranian language means store, hoard, store house … etc.,³² one of the clear indications of the fact that the compound name in Arabic Bayt al-Hikma for the royal libraries was established and used in new Persian. A quotation from the eleventh century by the Persian poet Nasir Khusraw (d. after 1070) lamented “the old good days in Balkh” and says, “it was the House of Wisdom (Khane-e Hikmet) alas it turned to ruins”.³³ This quotation clearly shows that the Arabic term was either used originally or as translated into Persian equivalent, “khane-e hikmet”.

Al-Nadim, our main source on the Abbasid Bayt al-Hikma and scholarly activities related to pre-Islamic sciences, does not use this term for the libraries of ancient Persians. He quoted the famous astronomer and astrologer, Abu Ma’shar Ibn Muhammed Ibn ‘Umar al-Balkhi (Albo Masar) (d. 886)³⁴ from his book on the variations of astronomical tables, on the eagerness of the Persian kings to collect and preserve books, but al-Nadim does not use any specific appellation for these libraries.³⁵ Al-Nadim, also quotes the words of Abu Sahl [al-Fadl] Ibn Nawbakht (d. 815) on Dahhak Ibn Qay, one of the legendary rulers of Persia who was overthrown by Faridun;

He [al-Dahhak] built a city in al-Sawad … He gathered into it the science of the scholars and built there twelve palaces, according to the number of the signs of the zodiac, calling them by the names [of these signs]. He stored the scholars’ books in them and caused the scholars themselves to live in them.³⁶

It is well known that both sources of al-Nadim were close associates of the Abbasid court and surely, they knew well what the House of Wisdom was. Abu Sahl, the famous physician and astronomer, was at the court of Harun al-Rashid and Abu Ma’shar was considered one of the most learned scholars of his time on the history of pre-Islamic Persia and had a strong sense of the intellectual primacy of Iran among the nations of Eurasian continent. He was also the munajjim (astronomer-astrologer) of al-Muwaffaq (r. 875–91), brother of the Caliph al-Mu’tamid (870–892).

It is clear from the above quotation and similar examples that this term was not used specifically for Iranian libraries that could have easily competed with the Abbasid Bayt al-Hikma as far as their volume and collections are considered.

It is quite interesting that al-Nadim, who does not use the appellation House of Wisdom for ancient Persian Royal libraries while providing information about alchemists and the art of alchemy in ancient Egypt, uses the appellation Bayt al-Hikma in a different context. In the tenth and at the end of the last chapter of his al-Fihrist, it is stated:

there were authors and learned men in this field among the people of Egypt where there was the beginning of talk about the Art and from which place, they derived it. The [well] known Barabi³⁷ which were the houses of wisdom and Mariyāh were in the land of Egypt.³⁸

Meanwhile, it seems that as an expression denoting the royal libraries, the appellation Bayt al-Hikma was not universally endorsed by Arab authors of the period. For example, Ibn Juljul (944–994), the Andalusian contemporary of al-Nadim, in giving information about Abbasid physicians and philosophers in his biographical book, does not use this term but settles for “Khazain al-Kutup” for the libraries.³⁹ In the thirteenth century, al-Qıftî uses the term khizane/khaza’in more frequently for the royal libraries, even for the great Alexandria library.⁴⁰ It is also noticeable that as time passed, its usage diminished and its Fatimid equivalent “Dar al-Hikma” came into use, as can be seen in the above example of al-Maqrizi in the fourteenth century.

From all the above examples and related discussions, it may be deduced that the term Bayt al-Hikma was not transferred from Greek and Persian but came out as a specific term for the royal library when it became larger, and when its contents were dominated by Hikma/Wisdom books. Contrary to the assumption of Eche, it did not appear in the Muawiya period and then disappear,⁴¹ nor was it the Arabic translation of the name given to old Sassanid royal libraries as argued by D. Gutas. As we understand, the compound name Bayt al-Hikma was not an extant, generic name, but it was composed in the Abbasid period and used to identify the collections in the great library joined to the palace during the periods of Harun al-Rashid and al-Ma’mun. Here an extensive collection of books related to pre-Islamic sciences and their catalogues were kept in a large area consisting of more than one room where scholarly activities related to these sciences could have been conducted. The plausible influence of Sassanid court libraries on their Abbasid counterparts cannot be disclaimed offhand; however, it may be explained within the wider context of the Persian influence on the Abbasid institutions. We believe the primary sources at hand do not lend themselves to any assumption to determine this influence. However, we are now better informed about Sassanid royal libraries as will be seen in Part VI.

Notes

1. Sa’id al-Andalusi, Science in the Medieval World, “Book of Categories on Nations”, trans. and edited by Semaan I. Salem and Alok Kumar, Austin, University of Texas Press, 1991, p. 43.

2. Katip Çelebi, Kashf al-Zunun, edited by E. Ihsanoglu, B.A. Marouf, Al-Furqan Edition, London, 2021, Vol. 3, pp. 412–415.

3. Al-Mas’udi, The Meadows of Gold, trans. and edited by Paul Lunde and Caroline Stone, London, Kegan Paul, 1989, p. 388.

4. AN I, p. 373.

5. AN I, p. 25.

6. AN I, p. 374.

7. AN II, pp. 142, 215.

8. AN I, p. 374.

9. AN I, p. 177.

10. AN I, pp. 25, 326, 374; AN II, pp. 142, 215, 326, 466.

11. AN I, pp. 24, 234, 373; AN II, p. 235.

12. AN I, pp. 13, 14, 44.

13. Y. Eche, pp. 44–48

14. Hunain Ibn Ishaq, Adab al-Falasife (Sentences des Philosophers), Abbreviation par Mohamed Bin Ali Bin Ibrahim, Edition Critique, note et Introduction par Abdurrahman Badawi, publication de l’Institue de Manuscript Arabes, Kuwait, 1985, p. 49.

15. Ibn Hawqal, Suret al-Ard (The face of the Earth), Leiden, 1938, 2nd Edition, Vol. I, s. 202.

16. E.W. Lane, Vol. I, p. 931.

17. AN II, pp. 157, 158, 172.

18. İsmail bin Ali b. Mahmud b. Muhammed bin Umar b. Shahinshah bin Ayyub, Abulfeda, Géographie D’aboulféda, Kitab Taqwim al-Buldun, M. Reinaud and M. Le bon Mac Guckin de Slane (eds.), Paris, L’imprimerie Royale, 1840, pp. 210, 216.

19. Here De Sacy’s edition of Kalila and Dimna will be used as it has the two introductions; however, it is not the best edition and is based on an inferior manuscript. Silvestre De Sacy, Calila et Dimna ou Fables De Bidpai, Paris, De L’imprimerie Royale, 1816.

20. De Sacy, pp. 33, 34.

21. On the introduction of Ali bin al-Shah al-Farisi; see De Sacy, p. 15 f.n. 1; Theodor Nöldeke, Die Erzahlun von Mausekönig und seinen Ministern, Göttingen, Dieterich’sche Verlags-Buchhandlung, 1879, pp. 6, 7; A.F.L. Beeston, “The ‘Ali Ibn Shah’ Preface to Kalilah wa Dimnah”, Oriens Journal, Vol. VII, 1954, p. 81; Carl Brockelmann, “Kalīla wa Dimna”, EI², IV, 1978, pp. 503–506.

22. De Sacy, pp. 25–31.

23. D. Gutas, pp. 54, 55.

24. Ibid.

25. Hamza Al-Isfahani, Kitab al-Amthal al-Sadira ‘an Buyut al-shi’r, Berlin manuscript, or. quart. n. 1215, ff.1a-6b.

26. Gutas, p. 55, fn. 45; Gregor Schoeler, Arabische Handschriften, Tell II, Stutgart F. Steiner, 1990, p. 308.

27. The Arabic text of al-Isfahani’s book was first published with the Latin translation made by J.M.E. Gottwaldt (ed.) Hamzae Ispahanensis, Kitab Tarikh Sinîy Muluk al-Ard wa al-Anbiya, Annalium, Libri X, Lipsiae, 1848, pp. 8, 9. For al-Isfahani, see Rosenfeld-Ihsanoglu, No. 196.

28. Personal communication with the Prof. Touraj Danyaee from California State University, Fullerton, 2003.

29. Kevin Bladel, “Bayt al-Hikma”, Encyclopedia of Islam, Third Edition, Leiden, Brill, 2009, s. 133–137.

30. Mansour Shaki, “The Denkard Account of History of the Zoroastrrian Scriptures”, Archiv Orientalni, Praha, Vol. 49, 1981, pp. 114–125.

31. D. Gutas, p. 36.

32. F. Steingass, Persian – English Dictionary, 6th Edition, Routledge, s. 1098.

33. Mahdi Muhakqiq, “Risalet Hunayn Ibn Ishaq”, in (the Book of collection of Papers delivered on the 1200th Anniversary of the Foundation of Bayt al-Hikma) Bayt al-Hikma al-Abbasi: A’rakat al-Madi ve Ru’yat al-Hadır, Baghdad, Bayt al-Hikma Institution Publishing, Vol. 2, 2001, p. 176.

34. Rosenfeld-Ihsanoglu, No. 88.

35. AN II, pp. 135–137; B. Dodge, pp. 576–678.

36. AN II, pp. 131–135; B. Dodge, p. 573.

37. Barabi plural Barba in Arabic, meaning old temples and monuments, see Gaston Wiet, “Barbā”, in Encyclopedia of Islam, Second Edition, Leiden, Brill, 2012.

38. AN II, p. 466; Dodge, p. 868, in the Arabic text of al-Nadim, her refers to the Egyptian Barabi as Buyut al-Hikma (houses of Wisdom); however, Dodge translate the term as the houses of learning.

39. Ibn Juljul, p. 61.

40. al-Qıfti, Leipzig, p. 355.

41. Y. Eche, p. 17.