1Pairing specific wines with particular foods, such as Vouvray sec with whitefish in beurre blanc sauce, is canonical among gourmets today (Larousse du vin 2004: 92), to the point that French gastronomy has elevated it to the level of an art (Pitte 2017). What might be less known is that it was also practiced in early Arabic culinary culture. Scenes from around the table in medieval Baghdad, described in poems and essays, reveal that wine was not consumed solely after meals, and that Arabs appreciated drinking wine with food. In his Fuṣūl al-tamāthīl fī tabāshīr al-surūr (Passages of poetic similes on joyful tidings) Abbasid prince ʿAbdallāh ibn al-Muʿtazz (d. 908) details, for example, food-drink pairing for a harmonious combination under the sub-title “taqdīr al-sharāb maʿa al-ṭaʿām wa-baʿdihi” (Appreciation of drinks with food and afterwards [after meal]) (Ibn al-Muʿtazz 1925: 66-67).
2This seems to contradict the prohibition of wine consumption in Islam that has become synonymous, along with the ban on eating pork, with Islamic food restrictions. But digging into both legal and historical sources shows that this prohibition was regularly contested by scholars, physicians and the consumers of these drinks, the latter including both members of the ruling elite and people from more ordinary walks of life.
3Drinking wine was part of a larger acculturation trend of pre-Islamic Sasanian court practices by the Abbasids. It is said, for example, that the Abbasid caliph al-Maʾmūn (r. 813-833) was aware that his drinking habits, of which he was so fond, were not in accord with the rules of Islam. Al-Masʿūdī (d. 956) tells how the tutor of caliph al-Rāḍī (r. 934-940) cited al-Maʾmūn who supposedly admitted:
I know that wine (zabīb) which I drink is forbidden
But I beg forgiveness of God, because he
Is kind and forgiving.
Those who drink grape liqueur (rubb)
Claiming that it is lawful and therefore
The wretched commit a double sin (al-Masʿūdī 2005 vol. 4: 269; Lunde and Stone 1989: 413).
- 1 Cookbooks were used by Lewicka in her description of Cairene foodways, quasi-alcoholic and alcoholi (...)
4This tension between the juridical prohibition on alcoholic beverages and the practice of drinking them described in classical Arabic literature and bacchic poetry (khamriyya) has been treated extensively by scholars (Kennedy 1997; Barbouchi 2009; Chebel 2004; Sadan 1977; Weber 2008: 60-67; Zakharia et al. 2014: 105-113). Although scenes of wine drinking are found in the sources, the drink’s composition as the main object is usually left out – except in recipes. This article proposes to delve into the question of what alcoholic drink was consumed: its name, ingredients, and production technique which determine, in practice, whether it was an alcoholic drink or not. All this information, described in food recipes, can be found in medieval Arabic cookbooks, sources little exploited to this day.1
- 2 See a detailed study of these subjects in my PhD dissertation: Yungman 2020.
5This article therefore explores the beverage recipes found in such cookbooks, and the ways in which they were incorporated into the textual Arabic culinary tradition. Seven Arabic cookbooks have survived from the Islamic East (Iraq, Syria and Egypt). They were written in court and urban milieus for the upper classes and diffused around the power centres of the Middle East in Abbasid Baghdad, Ayyubid Aleppo, and Mamluk Cairo for the purpose of documenting class practices and practical use for professional initiation or amateur cooks.2 Therefore, the examination of cookbooks of the entire medieval period (10th-15th centuries) makes it possible to evaluate whether and to what extent the composition of alcoholic beverages changed over that time. By using various sources, and both historic-anthropological and statistical methodologies, to analyse the proportion of alcoholic drink recipes, the article reveals the evolution of these drinks’ representation in the written heritage of the medieval Middle East up to the beginning of the early modern period.
- 3 The two remaining that do not include recipes for drinks are Kitāb al-Ṭabīkh (al-Baghdādī; 13th cen (...)
6In order to be licit to drink according to Islamic law, fermented beverages were permitted as long as they had a short fermentation, not exceeding three days – the presumed period after which the beverage allegedly becomes intoxicating (Abū Dāwūd, Sunan; Ibn Mājah, Sunan). Also, no fermentation agent (such as yeast or dregs) should be added, as stressed by al-Nasāʾī, according to whom “the dregs are what intoxicates (fī al-nabīdh khamruhu durdiyyuhu)” (al-Nasāʾī, Sunan). But as drink recipes show, this was not always the case. As a food category, the beverage (not only alcoholic) is found in five out of seven cookbooks: Kitāb al-Ṭabīkh (al-Warrāq; 10th century, Iraq), Kitāb al-Wuṣla ilā al-ḥabīb (anonymous; 13th century, Aleppo), Kitāb Waṣf al-aṭʿima al-muʿtāda (anonymous; 14th century, Egypt), Kanz al-fawāʾid (anonymous; 14th century, Egypt), and Zahr al-ḥadīqa (Ibn Mubārak Shāh; 15th century, Egypt).3 In each of these five books, this category is described in one or more dedicated chapters which group together a considerable number of recipes. But one cookbook from this collection is unique because its recipes explicitly mention alcoholic drinks.
7The earliest surviving Arabic cookbook from Iraq, Kitāb al-Ṭabīkh by one Ibn Sayyār al-Warrāq, composed of 132 chapters organised thematically by type of dish (615 recipes total) includes 16 recipes for explicitly alcoholic beverages in five chapters (Öhrnberg and Mroueh 1987: 298-303, 308-310; Nasrallah 2007: 454-463, 468-472) [tables 1-7].
Table 1: Recipes for alcoholic drinks in Kitāb al-Ṭabīkh, based on MS O
Chapter
|
Title
|
Title in Arabic
|
Explicitly intoxicating drinks*
|
Total no. of recipes
|
Chapter 113
|
Making excellent fuqqāʿ, plain and flavoured
|
ʿamal al-fuqqāʿ al-mukhtār al-sādhij wa-bi-l-abzār
|
4
|
8
|
Chapter 114
|
On humoral properties of wines and other beverages
|
fī ṭabʿ al-anbidha wa-l-ashriba al-muttakhadha
|
5
|
6
|
Chapter 119
|
On making wine from dādhī and the juice of grapes
|
fī ʿamal al-sharāb min al-dādhī wa-ʿaṣīr al-aʿnāb
|
4
|
4
|
Chapter 120
|
On making meads and raisin wines
|
fī ʿamal al-anbidha al-ʿasaliyya wa-l-zabībiyya
|
2
|
2
|
Chapter 121
|
On making wine from sugar and sugar-candy
|
fī ʿamal al-nabīdh min al-sukkar wa-l-fānīdh
|
1
|
3
|
Total
|
|
|
16
|
23
|
* determined by fermentation time or a fermentation agent
|
8Fuqqāʿ, as described in chapter 113, is a cereal-based beer-like beverage made from barley, wheat or bread. Its name, derived from the Arabic root f.q.ʿ, refers to the bubbles or foam created on its surface during preparation (Brinkmann 2014: 111), which included several steps, as detailed in this chapter’s second recipe. It was based on malted barley, used to create barley water, and sieved bread crumbs. Sugar and musk were added, and the brew was poured into big decanters in which beer had been previously brewed, but not more than ten days earlier, presumably a way of introducing yeast from an earlier batch. At this stage, the fuqqāʿ was put in the container overnight or from morning until night, then was replaced with a new batch, and this process would be repeated for ten days, until indication of fermentation (bubbles) appeared on the surface. Only then were flavouring herbs added: rue, tarragon, and mint.
9As barley-based beer recipes in this culinary manual show, the requirement of ten days’ fermentation was beyond the time permitted – and its result was undoubtedly intoxicating (Öhrnberg and Mroueh 1987: 298-299; Nasrallah 2007: 454-455). Moreover, since we do not know how long this drink was fermented in effect, and as in historical drinks the alcohol level is not certain (Nahoum-Grappe 1987: 439), it could be considered as simply “fermented.” In other sources it is also treated in an ambivalent way, both as alcoholic and non-alcoholic (Brinkmann 2014: 111), and, as stressed further by this chapter, there is an apparent textual evolution through the manuscripts. It appears in MS I 7322 as chapter 102 (on fuqqāʿ) that includes only two recipes, neither being for fuqqāʿ based on samīdh bread or giving indication of long fermentation; in fact, the recipe for honey fuqqāʿ calls for “drinking of it at the third day”.4
10This drink is widely present in the recipes of Kitāb al-Ṭabīkh and like other brewed drinks today, it has many varieties. Fuqqāʿ could be made from either bread, barley, rice, almond, or honey, and flavoured with spices, all based on barley water (māʾ al-shaʿīr). Unlike wine, that was drank during or after the meal, fuqqāʿ was recommended to be served before the meal, “yushrabu qabl al-ṭaʿām wa-lā yushrabu baʿdihi” (Ibn Qutayba 1925 vol. 3: 298). Furthermore, it had various digestive attributes; fuqqāʿ made of barley was even considered as beneficial against leprosy (judhām) (Ibn Qutayba 1925 vol. 3: 280).
11Considering the close connection of the Iraqi Abbasids with local Mesopotamian foodways, a taste for fermented cereal-based drinks was likely rooted in local practices. The same reason could explain the Egyptian preference for beer, mizr, as well as the cardinal place of fermented beverages in Egyptian cookery books, as will be discussed below. And indeed, Mesopotamian and ancient Egyptian beers are considered today to be the oldest documented, both archeologically and textually (Bottéro 2002: 141-147; McGovern 2009: 69, 241-250; McGovern 2014: 14; Lutz 1922: 72-96).
12Four more chapters in Kitāb al-Ṭabīkh (nos. 114, 119, 120, 121) contain the earliest recipes for alcoholic beverages in Arabic. They provide, with no apology, recipes for explicitly intoxicating drinks in a book that commences with the invocation Bismillāh (in the name of God), praises Almighty God and frequently ends recipes with the pious phrase Inshallāh (God willing). But these two were not always contradictory in Islamic history, and these expressions were also stylistic. Judging from their names, the drinks given in these chapters are various, for example in chapter 114: al-khamr min al-zabīb (wine from raisins) and sharāb muʿassal sādhij (plain honey wine), to name two [see full list in tables 3-7].
13Khamr, literally “wine,” is the term that clearly indicates an intoxicating drink and thus was forbidden legally to drink (Brinkmann 2014: 88-90). Its recipe goes as follows:
A recipe for making khamr from raisins: For each 10 raṭls of raisins, use 22 raṭls of sweet water. Follow the instructions we give in making al-khamr al-shamsī (wine fermented in the heat of the sun). Use the same ingredients required except for faṭṭāra, which you replace with honey. For each jar of wine, use 3 ūqiyyas of honey. Whisk the mixture with a miḍrab (beater) and stow the jars away in a sunny place, as we mention [in the last recipe] (Öhrnberg and Mroueh 1987: 302; Nasrallah 2007: 460).
A recipe for sun-fermented raisin wine, done the Egyptian way: For each 10 Egyptian raṭls [of raisins], use 27 raṭls of sweet water. Put aside 7 raṭls of the water mentioned and soak the raisins in the remaining 20 raṭls in leather bags smeared with pitch (maṭr muzaffat; the amphoras in which grape wine was aged were likewise smeared with pitch). Any other similar vessels will do. Let the raisins soak for six to seven days. Meanwhile, press and mash them very well with the hands two or three times a day. Empty the raisin mash into a large earthenware tub (mājūr), press it with your feet, and strain the juice. Next, thoroughly wash the dregs with the 7 raṭls of water you have set aside until nothing remains to be extracted from them. Strain the liquid [and add it to the rest of the juice]. Add to the raisin juice 1 fals fresh fennel seeds, 3 fals fresh wormwood, 8 dirhams laurel leaves, and 1 ūqiyya rose hips. Incorporate sugar to the must by whipping in 3 raṭls of excellent-quality yeast. Empty the must into jars and stow them away in a sunny place for two weeks. Adding wine dregs (durdī al-khamra), as well, will make even better wine (Öhrnberg and Mroueh 1987: 303; Nasrallah 2007: 462-463).
14Between these two recipes were four more, unrelated recipes. Clearly, one of these recipes is in fact relying on another, without which one cannot prepare it. This is a common practice in the non-“self-contained” instructional form of recipe (Tomlinson, 1986). Especially in historic recipes, it was a means to save space in an expensive-to-produce manuscript and to avoid redundant repetitions.
- 5 MS O: fol. 151v.
- 6 MS I 7322: fols. 102v-103v; MS N: fol. 161r.
15It is worth noting that although chapter 114 appears in all four manuscripts, the recipes for khamr, the most explicitly alcoholic beverage in name, and for zabīb shamsī occur in only one (undated) manuscript of Kitāb al-Ṭabīkh 5; in all three others these recipes are omitted. The recipes for the more debated zabībī muʿālij and nabīdh Miṣrī are given in all manuscripts; in MS I 7322 its equivalent is chapter 103 (on anbidha)6.
- 7 The practice of adding honey to grape wine existed already in ancient times. Long 2017: 81.
16Raisins are khamr’s most striking feature, compared to modern winemaking. Like the wine described by Mago in his treatise on Carthaginian viticulture (3dr-2nd c. BC) (McGovern 2009: 195), and by Pliny the Elder (Book XIV Chapter 11), four Arabic wine recipes (in chapters 114 and 120) are made from raisins, rather than fresh grapes. Considering the already high sugar content of raisins and the addition of honey or sugar,7 this recipe would result in the sort of wine containing over 15% alcohol known as straw or raisin wine (Lebègue 2012: 229-230), that would be categorised today as vin liquoreux.
- 8 Experimental archaeology within bioarchaeology is a growing field enabling scholars to recreate, am (...)
17The use of raisins could be for various practical and gustatory reasons. Practically, making wine from raisins allowed the winemaker to avoid relying on the supply of perishable fresh grapes. Raisins thus ensured the ability to produce wine all year long, regardless of climate conditions from drought to flood. In terms of taste, this method produced a wine agreeable to the sweet tooth. The grapes might have been dried on the vines or harvested and dried on straw mats under the sun (Millon 2013: 45), but the Arabic recipes do not specify a particular method. The technique of production indicated in the recipe, namely soaking raisins, then pressing and mashing them with the feet, then straining, adding yeast, and setting them aside for fermentation, resembles ancient winemaking procedures attested in bio-archaeological, iconographic and textual sources (McGovern 2003: 299-315; McGovern 2018; Ben-Gedalya 2018).8
- 9 Other types of fermented drinks, such as mizr, bitʿ or faḍīkh are not mentioned in the cookbooks an (...)
18Besides khamr, which was unanimously considered intoxicating and forbidden from early on, nabīdh was also subject to juridical debate. Its numerous recipes in Kitāb al-Ṭabīkh (chapters 114, 119, 120), showing diverse compositions, could explain this debate from a culinary point of view. Defined as “a comprehensive designation for intoxicating drinks” (Heine IE2), nabīdh (pl. anbidha) constitutes in fact a category in itself and encompasses various kinds of anbidha, depending on the main ingredient: grapes or raisin (nabīdh zabībī), date (al-nabīdh al-tamrī) or honey (nabīdh al-ʿasal) – all found in recipes. However, the mixture of fruits in the same drink is forbidden (al-Bukhārī, Ṣaḥīḥ; al-Kulaynī, Kāfī).9
19Controversy regarding nabīdh’s consumption preoccupied scholars such as the Iraqi encyclopaedist and littérateur al-Jāḥiẓ (d. 868), who was such an advocate of nabīdh that he wrote not one but two essays, “Risāla fī al-shārib wa-l-mashrūb” (Essay on the drinker and what is drunk) and “Fī madḥ al-nabīdh wa-ṣifat aṣḥābihi” (In praise of wine and the description of its properties), praising the consumption of nabīdh and arguing that it should be permitted (al-Jāḥiẓ 1964 vol. 3: 111-128; al-Jāḥiẓ 1964 vol. 4: 261-281). Al-Jāḥiẓ was well aware of winemaking techniques and the importance of some ingredients, such as dregs, in determining its lawfulness (al-Jāḥiẓ 1964 vol. 4).
- 10 MS I 7322: fol. 285r.
- 11 MS N: fol. 180v.
20This debate followed contemporary trends in including or excluding wine recipes in the manuscripts [table 2]. In the two latest manuscripts, one dated to 696/129710 and the other to 14 Shaʿbān 888/17 September 148311, chapters 119-121 are not for drinks; chapter 121 in MS O is found in chapter 115 in MS N and its only long-fermentation element (one month) is omitted. This late copy, and possibly the fact that MS N was copied in the holy city of Mecca, could be the reasons for the non-inclusion of these recipes and of chapters whose names indicate their alcohol content. On the other hand, the inclusion of nabīdh recipes can attest to the continuous debate on this drink into the 15th century. In this way these recipes were political tools.
Table 2. Evolution of chapters in Kitāb al-Ṭabīkh
|
MS O
|
MS H
|
MS I 732212
|
MS N
|
Chapter 113
|
Included
|
Included
|
= chapter 102
Only two recipes
|
Included
|
Chapter 114
|
Included
|
No recipes for khamr or zabīb al-shamsī
|
= chapter 103
Only three recipes, no recipes for khamr or zabīb al-shamsī
|
Only three recipes, no recipes for khamr or zabīb al-shamsī
|
Chapter 119
|
Included
|
Included
|
On making ushnān (potash) (equivalent to chapter 129 in MS O)
|
On making conserves (murabbayāt) and electuaries (juwārishnāt) (equivalent to chapter 125 in MS O)
|
Chapter 120
|
Included
|
Included
|
Decorum of handwashing before and after the meal (equivalent to chapter 130 in MS O)
|
On making condensed juices (arbāb) from tree crops (equivalent to chapter 126 in MS O)
|
Chapter 121
|
Included
|
Included
|
Table manners when eating with dignitaries (equivalent to chapter 131 in MS O)
|
= chapter 115
The indication of using after a month in second recipe is replaced by “then use it” (fol. 162r)
|
21As the recipes in tables 3-7 show, the beverage’s name could be deceiving, and its potential alcohol content depended on the preparation technique:
Table 3. Chapter 113: Beer-like fuqqāʿ
Recipe
|
Starch source
|
Fermentation agent
|
Flavorings
|
Time
|
al-fuqqāʿ al-muttakhadh min al-khubz al-samīdh (beer made with bread of fine flour)
|
Malted barley
|
Bread
Sugar
|
(salt)
Rue
Tarragon
Mint
Musk
|
10 days
|
fuqqāʿ al-aruzz (rice beer)
|
Rice
Barley water
|
Sugar
|
Rue
Tarragon
Mint
|
10 days
|
al-fuqqāʿ al-muttakhadh bi-l-lawz (almond beer)
|
Almonds
Barley water
|
Sugar
|
Rue
Tarragon
Mint
|
10 days
|
al-fuqqāʿ al-Qurashī al-muttakhadh bi-l-abzār (Qurashī beer flavored with spices)
|
Malted barley
Barley water
|
Sugar
|
Pepper
Cassia
Spikenard
Cloves
Nutmeg
Mace
Black cardamom
Chestnut
Long pepper
Rue
Tarragon
Mint
|
10 days
|
Table 4. Chapter 114: Raisin wines and meads
- 13 This plant’s identification is based on the term used, murratayn akhḍar, explained in Lisān al-ʿAra (...)
Recipe
|
Main ingredient
|
Fermentation agent
|
Flavorings
|
Sun
|
Time
|
al-khamr min al-zabīb (wine from raisins)
|
Raisins
|
Honey
Sugar
(optional wine dregs, durdī al-khamra)
|
Fennel seeds
Wormwood
alāʾ13
Laurel leaves
Rose hips
|
Yes
|
2 weeks
|
sharāb muʿassal sādhij (plain honey wine)
|
Honey
|
Honey
|
Saffron
|
NA
|
40 days
|
al-nabīdh al-zabībī al-muʿālij (curative raisin wine)
|
Raisins
|
Honey
|
Spices are optional
|
NA
|
NA
|
nabīdh Miṣrī (Egyptian wine)
|
Honey
|
Honey
|
No
|
No
|
40 days +
4 months
|
zabīb shamsī (sun-fermented raisin wine)
|
Raisins
|
Sugar
Fermented dough (faṭṭāra ʿāl)
(optional wine dregs, durdī al-khamra)
|
Fennel seeds
Wormwood
alāʾ
Laurel leaves
Rose hips
|
Yes
|
2 weeks
|
Table 5. Chapter 119: Date liquors and grape wines
Recipe
|
Main ingredient
|
Fermentation agent
|
Flavorings
|
Sun
|
Time
|
nabīdh al-tamr wa-l-dādhī (wine from dates and dādhī)
|
Dates
|
dādhī – hop cones (?)
Honey
|
No
|
NA
|
3 days preparation + 3 months storage
|
nabīdh min dibs wa-dādhī (wine from molasses and dādhī)
|
Date syrup
|
dādhī – hop cones (?)
Honey
Natural yeast: whipping mixture for three days before sealing vessels
|
No
|
Yes
|
20 days +
2 months
|
ʿaṣīr al-ʿinab al-shamsī (sun-fermented grape juice)
|
Grapes
|
Natural yeast (grape leaves cover in first step)
|
Vessels are fumigated with mastic and aloe wood
|
Yes
|
5 days preparation + 40 days
|
ʿaṣīr al-ʿinab (grape juice – wine)
|
Grapes
|
Natural yeast (grape leaves cover in first step)
|
No
|
NA
|
5 days preparation + one year
|
Table 6. Chapter 120: Raisin wine and mead
Recipe
|
Main ingredient
|
Fermentation agent
|
Flavorings
|
Sun
|
Time
|
nabīdh zabībī (raisin wine)
|
Raisins
|
Honey
|
Saffron
Spikenard
Mastic
Musk
|
No
|
40 days
|
nabīdh ʿaslī bi-ghayr zabīb (mead without raisins)
|
Honey
|
Natural yeast (sealed after 50 days)
|
Saffron
Mastic
|
No
|
Over 50 days
|
Table 7. Chapter 121: Sugar and sugar-candy liquors
Recipe
|
Main ingredient
|
Fermentation agent
|
Flavorings
|
Sun
|
Time
|
sharāb al-sukkar wa-l-fānīdh (sugar wine)
|
Sugar
|
NA
|
Aloe wood
Musk (in winter)
Camphor (in summer)
|
NA
|
NA
|
sharāb al-fānīdh al-Khurāsānī wa-l-fānīdh al-Shajarī (wine from chewy candy)
|
Sugar candies
|
NA
|
Saffron
Spikenard
Cloves
Cassia
Nutmeg
|
NA
|
1 month
|
sharāb ākhar min al-sukkar (another sugar wine)
|
Sugar
|
NA
|
Saffron (for color)
|
NA
|
NA
|
22The 16 explicitly intoxicating beverages in these five chapters of Kitāb al-Ṭabīkh, following MS O, are classified by main ingredient: cereal-based beer-like fuqqāʿ, raisin wines and meads in two distinct chapters (114 and 120), date liquors and grape wines, and sugar-based liquors. They were mostly flavoured with various aromatic substances, spices and perfumes, besides some recipes for meads, date liquors and one grape wine. Considering saffron’s main trait as a colouring rather than a flavouring, its addition suggests that these wines or liquors were clear (yellow) rather than red in colour.
- 14 Arabs distinguished wine colours even before Islam. Maraqten 1993: 102. An equivalent approach is a (...)
23Even though the recipes do not specify details of the resulting wine, its category, colour, or taste, other sources do provide this type of information. Ibn al-Muʿtazz divides the colours of wine into four: red, yellow, white, and black (qismat alwān al-sharāb wa-hiya arbaʿa: al-aḥmar, wa-l-aṣfar, wa-l-abyaḍ wa-l-aswad) (Ibn al-Muʿtazz 1925: 26). He also categorises wine into three stages of maturation: old wine (nabīdh ʿatīq), “medium” wine (khamr mutawassiṭ) and new wine (khamr jadīd) (Ibn al-Muʿtazz 1925: 23-26). All these descriptions complement the recipes,14 like the one for sun-fermented grape wine that says: “the longer this wine is allowed to age, the better it gets in aroma, taste, and colour” (Öhrnberg and Mroueh 1987: 308; Nasrallah 2007: 469), even though it does not specify what that colour would be.
- 15 As discussed extensively by Nasrallah and Brinkmann, dādhī (also: dādī/dhādhī/dāzī) the “mysterious (...)
24In order to initiate the fermentation process, most fuqqāʿ recipes rely on sugar; wines and non-grape liquors use either honey or sugar or both. Date liquor had a single substance added, dādhī, whose exact botanical identification is not certain, but it has been proposed as hop cones (Nasrallah 2007: 545-551).15 Otherwise, grape wine used natural yeast, obtained by leaving a hole in middle of grape leaves covering the vessels’ opening. The recipe for khamr is unique in this respect for being the only one to instruct the maker to add wine dregs; but this fermentation substance will reappear in later cookbooks’ drink recipes. Only three recipes include sun-fermentation, using the adjective shamsī (from shams for sun). Also, all recipes share one common feature: none of these drinks is prepared using distillation; they are all fermented, either by sun-fermentation or by long fermentation. That is despite the fact that the technique of distillation was known and used in the Islamic Middle East for perfumes and beverages as early as the 8th century (al-Hassan 2009; Perry 2017: 262-277), and described in drink recipes for the first time in later cookery books.
25Lastly, in 14 recipes the fermentation time exceeds the permitted three days, extending from ten days (fuqqāʿ) or two weeks (khamr shamsī and zabīb shamsī) to one month, 40 or 50 days, two or four months – up to a year in the case of ʿaṣīr al-ʿinab (grape wine) in chapter 119. This means that these drinks did not figure in the debate because they were explicitly and openly intoxicating. In other cases, the short fermentation or lack of its mention altogether leaves the result as an open question.
26 The existence of recipes for intoxicating beverages in this specific cookbook, Kitāb al-Ṭabīkh, is no accident. In chapter 115, following Hippocratic-Galenic medical theory, al-Warrāq considers wine’s medicinal benefits. According to this chapter, sun-fermented wine prepared from honey and raisins was recommended for people who had phlegmatic humours who should even “have a lot of it” (Öhrnberg and Mroueh 1987: 304). Other textual clues, such as names of Abbasid caliphs, indicate that it was compiled in a court milieu. Members of the Abbasid court began producing private recipe collections as early as the 8th century (Yungman 2020), during what is considered as the Abbasid “Golden Age.” This pastime activity reflects the Abbasids’ fascination with Greek sciences and bibliophilia (Gutas 1998; Touati 2003).
- 16 The lack of textual evidence (recipes) from the Umayyad period does not permit us to know the compo (...)
27In the process of creating their haute cuisine, and laying the foundation for Arabic culinary traditions, the Abbasids chose the Sasanian court and Nabatean cuisine as primary models (Waines 2003: 574; Waines 2002: xxxii-xxxiii; Nasrallah 2007: 45-55).16 They did so because of the absence of a structured institutional cuisine during the reigns of the founders of Islam, namely the prophet Muḥammad and the first four caliphs, and because of the Persian Shuʿūbiyya movement’s influence. Thus, like other pre-Islamic neighbouring royal cuisines and local pre-Islamic Iraqi views on wine, which were largely positive, Kitāb al-Ṭabīkh and its alcoholic drink recipes became a tool in this process of institutionalisation of caliphal cuisine and post-Abbasid court practices in general.
28Furthermore, the presence of local communities from different cultural and religious backgrounds, and notably Christians and Jews, who have been consuming fermented drinks since antiquity and to whom the consumption of fermented drinks was not forbidden, influenced the spread of intoxicating drinks amongst the ruling Muslim elites. Thus the many Christian monasteries in the Middle East, described in detail in al-Shābushtī’s (d. 1000) Kitāb al-Diyārāt (Book of Monasteries), were often visited by dignitaries, especially Abbasid caliphs, who took up residence there to drink, listen to music and enjoy the beautiful scenery and surrounding vineyards (al-Shābushtī 1986). It is therefore not surprising that a sophisticated culture of wine making developed all around the Middle East.
29At first glance, it might seem that only Kitāb al-Ṭabīkh contains recipes for explicitly intoxicating drinks. However, attentive examination of recipes from later cookery manuals indicates that recipes for fermented beverages continued to be included; these drinks may have had other names, but their technique would have produced a drink that could have been intoxicating.
30Through a broader perspective on the recipes, we may now examine the place of fermented drinks in three later cookbooks, all representing the upper-class urban cuisine. Their geographical reach went beyond Iraq to Syria and Egypt, thus showing continuity and adaptation to the social circumstances in which they were produced. By nature, they kept adding drink recipes and reinvented their range as well as composition, at the margin of licit drinking and at times crossing that line – though none give recipes for drinks explicitly deemed khamr or nabīdh. These fermented drinks, most of them cereal- or fruit-based, push the limits of “the rules of the game” of society at the time.
- 17 Charles Perry, the editor of the latest edition, found two textual families to which the manuscript (...)
31The 13th century Syrian cookbook Kitāb al-Wuṣla ilā al-ḥabīb (635 recipes, henceforth Wuṣla) includes in its chapter “On drinks” (fī al-ashriba) 19 recipes for variety of drinks, of which five are for beer-like beverages, made of cereals (wheat, flour, rice) and fermented. They were probably alcoholic, although at low percentages, because they are all based on fuqqāʿ17 and in all of them yeast is added (khamīrat al-fuqqāʿ) – as already explained, a forbidden ingredient in the production of licit beverages under Islamic law. Despite the fact that no recipe in this book specifies a long fermentation – some even say to ferment only “for half a day” (niṣf nahār, Perry 2017: 30) – they all mention that foaming should occur in order to know if the drink is ready to be consumed, using the verb ṭalaʿa, to rise.
32Besides the explicit reference to beverages prepared with yeast in Wuṣla, one particular recipe for a drink named ṣurmār is also intriguing. The recipe, found uniquely in Text Family B of this cookery book,18 says explicitly: “If you want it to be slightly intoxicating, put in pounded qamlaq (hops, Humulus lupulus), half a dirham for every mann of water; it will improve it (wa-in aradta an yakūn fīhi iskār fa-yujʿalu fīhi qamlaq madqūq, li-kull mann al-māʾ niṣf dirham, wa-huwa yujawwiduhu)” (Perry 2017b: 8-9).
33Another recipe is for aqsimā elixir (iksīr aqsimā), which could be the very first comestible recipe using a distillation technique, using the term qaṭr. It says:
Twenty-four dirhams of pepper, the same of ginger, four dirhams each of rosehips cleaned from their stalks, spikenard, betel nut, cardamom, clove, mace, and nutmeg, 200 dirhams of rose petals, and 1,200 dirhams of sharp vinegar. Pound the dry ingredients, soak them with the vinegar for two days, distil on a low fire, and set aside. When you want to make aqsimā, dissolve sugar, squeeze lemons into the syrup, and macerate a stalk of rue in it. Throw on a little of the distillate and it will immediately turn into aqsimā (Perry 2017b: 4-7).
34Aqsimā is an oxymel, a medicinal syrup of honey and vinegar (Bouras-Vallianatos 2021: 967), here flavoured with various spices and made with sugar rather than honey. Although it is not a recipe for a distilled beverage such as ʿaraq, but for a medicinal extract, it attests the use of a distillation technique, which probably exceeded medicinal and perfumery purposes and was used otherwise in producing beverages.
35In the most exhaustive medieval cookbook in Arabic extant today, Kanz al-fawāʾid (14th century, Egypt; 830 recipes, henceforth Kanz), chapter 12, “fī ʿamal al-fuqqāʿ wa-ghayruhu” (On making fuqqāʿ and other drinks), contains 43 recipes for fermented beverages. It is divided into two sub-sections, “Chapter on making aqsimā, fuqqāʿ, and māʾ al-shaʿīr (malted barley drinks),” and “Chapter” (bāb) (Marín and Waines 1993: 152, 153; Nasrallah 2018: 276, 277) that adds more recipes for fermented drinks. This suggests that these sections were copied from another source (or sources), perhaps ones entirely dedicated to beverages, lost today (Nasrallah 2018: 276 no. 31, 277 no. 40). The inclusion of aqsimā recipes in this chapter, using a distilled substance as in Wuṣla (the aqsimā elixir), could indicate a change in its categorisation. A closer analysis of Kanz’s recipes for fermented drinks offers these findings:
Table 8. Fermented beverages in Kanz
Type of drink
|
Number of appearances
|
Recipes with possible alcoholic content
|
fuqqāʿ
|
19
|
1
|
aqsimā
|
7
|
3
|
sūbiyya
|
6
|
1
|
shishsh
|
4
|
2
|
khamīrat al-aqsimā/ḍārī
|
4
|
–
|
māʾ al-shaʿīr
|
2
|
–
|
Total
|
42
|
7
|
36The recipes for fermented beer-type drinks in Kanz demonstrate a shift in their composition and perhaps conception; most of the 31 recipes for fermented drinks specify only short fermentation periods, less than three days. In addition, only a handful, seven, are potentially alcoholic because they include some sort of beer yeast (khamīrat al-fuqqāʿ). Also, Kanz proposes three recipes for khamīrat aqsimā, a distilled aqsimā-base for making instant oxymels (Marín and Waines 1993: 151, 152, 158; Nasrallah 2018: 275, 276, 283 recipe nos. 402, 404 & 423). Two aqsimā recipes are clearly alcoholic, one is even distilled (recipe no. 420): “distil the liquid on a very low heat and store it away” (Marín and Waines 1993: 157; Nasrallah 2018: 281-282), the other uses the distilled base indicated in previous recipes, but its fermentation should last only overnight (recipe no. 424) (Marín and Waines 1993: 158; Nasrallah 2018: 284).
- 19 In his Zahr al-ḥadīqa, an abridged version of Kanz, Cairene Ibn Mubārak Shāh (d. 1458) includes a c (...)
37Nine others instruct the reader to pour the liquid into jars with traces of previous fermentation, viz. yeast spores (recipe nos. 396, 403, 410, 417, 419, 429, 430, 431, 434). For example, one recipe for a sort of rice beer called sūbiyya says to boil rice, then soak ʿajwa (date paste) overnight, boil and strain, then dissolve the rice in the date juice and boil them together until the mixture thickens. This is flavoured with spices and orange peel, then the liquid is left to ferment in a jar, “with traces of previous fermentation” (Marín and Waines 1993: 160; Nasrallah 2018: 286). This last instruction, “wa-yukhammar fī wiʿāʾ ḍārī,” implies fermentation that could result in an alcoholic liquid. The reason for this transformation in Kanz’s recipes is not certain but could be related to societal considerations under Mamluk rule, the period when this cookery book was written. It could also be simply the taste of the compiler for these specific beverages.19
- 20 MS I 62: 310; MS I 22: 276.
- 21 MS I 62: 49-50; MS I 22: 261-263.
38The anonymous author of Kitāb Waṣf al-aṭʿima al-muʿtāda (14th century, Egypt; 415 recipes, henceforth Waṣf) classifies all drink recipes in one chapter as “digestive” drinks (fī al-hāḍimāt) (Perry 2006: 440)20. Despite its neutral title, this chapter contains vaguely instructed beverages that could be fermented. Fuqqāʿ khāṣṣ (special fuqqāʿ) is prepared on the basis of biscuits (kaʿk) rather than bread, added to pomegranate juice, sugar, flavoured with the spice blend aṭrāf al-ṭīb, pepper, ginger, and scented with musk and rose water. What could have made it intoxicating is the fact it is not cooked, and no instruction is given to boil it or reduce its quantity by two-third, in addition to the imprecise direction to put the final resulting liquid “in the usual small jugs (kīzān). Then you seal every jug (kūz) with the remaining rose-water and musk. And it is left in it a little (qalīl).” (Perry 2006: 441)21. How long a “little” would be, is not detailed. A similar recipe found in Text Family B in Wuṣla uncovers an interesting instruction; in Waṣf the recipe in the earlier MS I 22 (copied in 1373) adds “empty,” (p. 263) in Wuṣla it clarifies the instruction by saying “do not fill the jugs but leave some headroom” (Perry 2017b: 8-9), implying that the fermentation process will create foam, which requires space inside the jug.
- 22 MS I 62 : 49; MS I 22: 261-263.
39Contrary to this ambiguous instruction of “a little while,” Waṣf contains another, much bolder instruction that must have resulted in an alcoholic beverage. This time, the recipe has a vague title, sharāb jalīl al-qadr (drink of sublime power) that does not explicitly call the drink “wine” but from its composition and preparation method it was clearly some kind of raisin wine or raisin alcoholic liquor. Despite the fact that the first phase entails soaking raisins in water for not over three days (in winter) and a day and a night in summer, as required by law and specified repeatedly in ḥadīth collections, the later addition of honey, which could raise its alcohol content, and the striking instruction to leave it “in its vessel, sealed, for three months” (Perry 2006: 441)22 could not have produced anything other than an alcoholic drink. Certainly, sharāb could be translated as either “drink” or “syrup” and it could be read as such because it is strained. But from other recipes it is obvious that there is no need to set the preparation aside for three months before using it as a mere syrup.
40The lack of recipes for explicitly intoxicating beverages specified by name in these culinary collections could be filled in by supplementary sources. Ibn Ḥawqal reports already in the 10th century that Egypt was widely known for two alcoholic beverages: mead (sharāb al-ʿasal) and mizr. He declares that Egyptian mead is “reputed in the entire world thanks to its delicacy, the quality of its fragrance and the intoxication caused by its flavour” (Ibn Ḥawqal 2001: 159). Mizr reached Sind and was considered there “a fermented drink of the Egyptians,” but he admits not knowing anything of it or its composition (Ibn Ḥawqal 2001: 317). It is also mentioned in the Cairo Geniza, a rich cache of manuscripts discovered in a synagogue in Old Cairo, dated to mid-10th to mid-13th centuries (Goitein 1983 vol. IV: 261).
- 23 On the geographical and social context of these documents see: Goitein 1967 vol. I: 16-28.
41In addition, two undated recipes for honey wine in Judeo-Arabic are found outside the cookbooks in fragments from the same Geniza, both detailing the ingredients and quantities. One is nabīdh tayyib, made of honey and spices (lichen, ginger, pepper, saffron). It is fermented under the sun for seven days, indicating clearly that it contained alcohol (JTS ENA 2808.22: 1r; Goitein 1983 vol. IV: 260). The second recipe is shorter, for nabīdh dādhī and also made of honey (CUL T-S Ar. 43.71: 1v; Goitein 1983 vol. IV: 260), but without dates as is described in the Iraqi recipe from Kitāb al-Ṭabīkh. The main challenge, as in any Geniza fragment, is to situate these recipes accurately; when, where, by whom, and why they were written, and in which social context?23 What is certain is that they echo a tradition of fermented beverage-making and a taste for these drinks that was shared by Muslims, Christians and Jews.
42The German Dominican friar Burchard of Mont Sion (d. 1285), who visited the Holy Land, reports on the fine wines produced by local Christians – but not only them – and provides an external viewpoint on the subject: “There are many vineyards in the Holy Land, and would be more, but the Saracens, who now own the land, do not drink wine (except some of them, secretly) and so do not cultivate vineyards but destroy them, except for a few Saracens who happen to live nearer the Christians and cultivate them for profit to sell grapes and wine to the Christians” (Burchard of Mont Sion 2019: 186-189).
43Similarly in Egypt, based on testimonies from the Geniza, S. D. Goitein argues that “in our period […] a certain laxity prevailed” [concerning wine and alcohol drinking by Muslims] (Goitein 1967 vol. I: 122-123). Furthermore, the Geniza documents show that “the huge sales of wine referred to could hardly have been destined for the Christian and Jewish population alone, especially as many people, and certainly not only Jews, used to buy grapes and have them pressed and made into wine for home consumption” (Goitein 1967 vol. I: 123). The existence of wine recipes further supports these arguments.
- 24 More on the shape and use of this vessel see Lewicka 2011: 466-467.
44A brief analysis of drink types in the three cookbooks studied (Wuṣla, Waṣf and Kanz) shows that the drink category most represented in the recipes is fermented, beer-like drinks, with 39 recipes. In fact, fuqqāʿ was so popular that its recipient vessel was named after it (Goitein 1983 vol IV: 261).24 One reason for their popularity is their main ingredient, barley and wheat, the prevalent crops of the region. Drinks were then another manifestation of these cereals in the Arab bread-based culture. It is likely that grain drinks made from these recipes did indeed contain alcohol, even if at low percentages. In an experiment conducted by chemist Alfred Lucas in the 1920s on beer samples collected from Cairo’s market, he found that these beers contained 6.2-8.1% alcohol (McGovern 2009: 247). In the 1970s, Sabry Morcos and his colleagues tested the cereal-based beverage būẓa as sold in Cairo and found that it showed an alcohol content of 3.8% when fermented for one day and 4.45% at three days (Morcos et al. 1973: 1158).
- 25 The same word has made its way to modern Hebrew as ḥanṭarīsh meaning an unreliable person who speak (...)
45Even though cookbooks no longer detailed winemaking in recipes, wine kept preoccupying Muslim scholars far into the 15th century. Egyptian poet and anthologist Muḥammad ibn Ḥasan al-Nawājī (d. 1455) describes in his Ḥalbat al-kumayt (The Racecourse of the bay), among other festive topics, the names of wines, and their origins, traits, and uses according to physicians (al-Nawājī 1938: 6-25). Al-Nawājī enumerates 36 synonyms for wine, some borrowed from different languages like khandarīs (old wine) coming from Persian (Steingass 1963: 476). Also, khand in Persian means “laughter” and rīsh “to show oneself stupid and inexperienced” (Steingass 1963: 603).25 Another synonym, qahwa, had received around at the same period, or later, its modern-day meaning of coffee.
- 26 Already during the prophet Muḥammad’s times there are ḥadīths testifying the uncertainty concerning (...)
46Without being able to know for certain what wine made from these recipes was like, the name of the drink might be the first indication of its presumed alcohol level content, and therefore the controversy should be seen in the light of these names.26 Despite the clear prohibition on renaming alcoholic drinks according to a ḥadīth that says: “It was narrated from ʿUbādah ibn Sāmit that the Messenger of Allah said: ‘People among my nation will drink wine, under some other name that they will give it’” (Ibn Mājah, Sunan), the recipes present in cookery books bear various names that sometimes disguise long-fermentation production techniques.
47If in Abbasid Iraq alcohol drinking by the elites was often not a secretive practice, in the Mamluk sultanate, which considered itself to be the “saviour of Islam” (Northrup 1998: 302), this had become an obvious contradiction with Islamic dietary laws, and as such had to be modified and adapted in order to be accommodated to its consumers. The creative solution was euphemism; the most brilliant example is sharāb jalīl al-qadr, which could not have raised much objection in its very name. Its content, obviously, was meant for a reader who was probably not only not shocked by it, but most likely requested its inclusion, because according to one manuscript’s colophon, it was copied for a certain Amīr Rashīd in Mamluk Egypt27 (Perry 2006: 450). Another example from the recipes is calling a grape wine aged one year as “ʿaṣīr al-ʿinab,” grape juice. Even al-Jāḥiẓ admits this practice when he says: “it sometimes happens that a thing classed as legally forbidden undergoes a sort of metamorphosis whereby the effect upon it of a new colour, smell, taste, etc. alters it in such a way that a semantic shift occurs and, lo and behold, what had been forbidden is transformed into something legitimate!” (al-Jāḥiẓ 1964 vol. 4: 274; Colville 2002: 140).
48This brings the issue of what was, indeed, forbidden to drink. Ḥanafī jurists, for example, distinguished between wine, khamr, explicitly condemned in the Qurʾān, and all other types of fermented beverages. In that way, they “permitted the limited consumption of some intoxicating substances” (Haider 2013: 49). In fact, this was their main disagreement with other schools of jurisprudence (Haider 2013). It could explain the laxity in including recipes for drinks that are not khamr, like nabīdh or fuqqāʿ, although they could have contained alcohol.
49The formation of legal rules concerning wine and alcoholic beverages during Abbasid times therefore revolved around the linguistic question posed by al-Jāḥiẓ as an example of a source of confusion:
If linguistic experts in bygone times had believed that the word “wine” covered all intoxicating beverages, there would have been no need for them to seek the opinion of ḥadīth scholars regarding the category in which any given drink fell. […] The scholars who disagree with my position on the legitimacy of nabīdh are contradicted by their own admission that, although the large number of intoxicating drinks in existence at the time were well-known by name, type, quality and origin, God singled out wine alone for prohibition – so leaving all the rest in the category of the permissible (al-Jāḥiẓ 1964 vol. 4: 275; Colville 2002: 140).
50Recipes for fermented drinks and wine help in clarifying the confusion raised in the debates about the nature of the alcoholic drink – from the point of view of its composition and production technique.
51Conclusion
52Arabic historical cookbooks, ascribed to a shared collection of recipes, testify to the geographical diffusion of textual culinary knowledge from Iraq-based Abbasid caliphate westwards to Syria and Egypt. Their recipes also mark a diachronic evolutional shift in the preparation of alcoholic drinks and their consumption related to the respective power centre. Recipes for alcoholic drinks in Arabic cookbooks show that at least in medieval Islam, alcohol prohibition was not unanimously and instantly accepted but was constantly challenged. That was probably because of these drinks’ social, medicinal and ceremonial roles in the Middle East in the millennia before Islam (Maraqten 1993; Brinkmann 2014: 83-87).
53Understanding the prohibition on alcoholic drink should be then nuanced and reconsidered. Certainly, the five main ingredients (grapes/raisins, date, wheat, barley, honey) for potentially alcoholic drinks, repeatedly stressed in the ḥadīth collections (Brinkmann 2014: 89-90), are all found in the recipes. However, the latter show that what made the difference between a licit and a forbidden drink was not the main ingredient, but the technique (fermentation time, storage conditions, type of vessels) as well as the existence of a fermenting agent. That is because ḥalāl (licit) drinks were made from the same ḥarām (forbidden) ingredients: sawīq, a refreshing drink made of toasted barley, grape juice, and other fruit juices, was allowed and consumed. Alcoholic beverages did exist, and these recipes are the bridge to their historic existence. In their sometimes ambiguous instructions, especially concerning the crucial period of fermentation, the recipes could have been interpreted in many ways, creating, without doubt, licit syrups and refreshing drinks, but also fermented alcoholic ones.
54As in other cultures (Bianquis 2012: 22-23), the alcoholic drink in pre-modern Islamic societies was a “boundary object” because it was constantly on the border between the licit and the illicit. It could be seen as an object on a “fracture line” that distinguished between pre-Islamic (and openly wine imbibing) and “sober” Islamic societies. But as the recipes show, this line never really existed; there was no fraction but continuity, especially considering that eating and drinking practices are continually changing in the “long conjuncture” (Braudel 1961: 723-728). In fact, in terms of drinking habits, the 15th and 17th centuries mark a shift in the advent of coffee and tea that “replaced” wine and alcohol in the Middle East, especially in hospitality (Aubaile-Sallenave 2005: 189-190). But the story of wine and alcohol in Islam is far from ending in medieval times.