Харджа: Финальный рефрен андалузской поэзии мувашшах

A kharja or kharjah (خرجة ; lang es ˈxaɾtʃa; carja ˈkaɾʒɐ; also known as markaz), is the final refrain of a muwashshah (مُوَشَّح 'girdle'), a lyric genre of al Andalus (the Iberian Peninsula under Muslim control) written in Arabic or Andalusi Romance ("Mozárabic"). The muwashshah consists of five stanzas (bait) of four to six lines, alternating with five or six refrains (qufl); each refrain has the same rhyme and metre, whereas each stanza has only the same metre. The kharja appears often to have been composed independently of the muwashshah in which it is found.

About a third of extant kharjas are written in Classical Arabic. Most of the remainder are in Andalusi Arabic, but there are about seventy examples that are written either in Iberian Romance languages or with significant Romance elements. None are recorded in Hebrew, even when the muwashshah itself is in Hebrew. Generally, though not always, the kharja is presented as a quotation from a speaker who is introduced in the preceding stanza. It is not uncommon to find the same kharja attached to several different muwashshahat. The Egyptian writer Ibn Sanā' al Mulk (1155–1211), in his Dar al Tirāz (a study of the muwashshahat, including an anthology) states that the kharja was the most important part of the poem, that the poets generated the muwashshah from the kharja, and that consequently it was considered better to borrow a good kharja than compose a bad one. Kharjas may describe love, praise, the pleasures of drinking, but also ascetism.

Of the approximately 600 known secular Arabic muwaššaḥāt, there are almost 300 kharjas in vernacular Andalusi Arabic and over 200 in Standard Arabic (فُصْحَى), though some of the vernacular kharjas are essentially Standard Arabic with a vulgar gloss. About 50 are in Andalusi Romance or contain some Romance words or elements.

Since the kharja may be written separately from the muwashshah, many scholars have speculated that the Romance kharjas were originally popular Spanish lyrics that the court poets incorporated into their poems. Some similarities have been claimed with other early Romance lyrics in theme, metre, and idiom. Arabic writers from the Middle East or North Africa like Ahmad al Tifashi (1184–1253) referred to "songs in the Christian style" sung in al Andalus from ancient times that some have identified as the kharjas. Other scholars dispute such claims, arguing that the kharjas stand firmly within the Arabic tradition with little or no Romance input at all, and the apparent similarities only arise because the kharjas discuss themes that are universal in human literature anyway.

Modern translations of the Romance kharjas are a matter of debate particularly because the Arabic script does not include vowels. Most of them were copied by scribes who probably did not understand the language they were recording, which may have caused transmission errors. A large spectrum of translations is possible given the ambiguity created by the missing vowels and potentially erroneous consonants. Because of this, most translations of these texts will be disputed by some. Severe criticism has been made of García Gómez's editions because of his palaeographical errors. Further debate arises around the mixed vocabulary used by the authors. Most of the Romance kharjas are not written entirely in Romance, but include Arabic elements to a greater or lesser extent. It has been argued that such blending cannot possibly represent the natural speech patterns of the Romance speakers, and that the Romance kharjas must therefore be regarded as macaronic literature. A minority of scholars, such as Richard Hitchcock contend that the Romance Kharjas are, in fact, not predominantly in a Romance language at all, but rather an extremely colloquial Arabic idiom bearing marked influence from the local Romance varieties. Such scholars accuse the academic majority of misreading the ambiguous script in untenable or questionable ways and ignoring contemporary Arab accounts of how Muwashshahat and Kharjas were composed.

Ibn Sanāʾ al Mulk, a 12th century Egyptian poet, wrote an anthology and study of the muwaššaḥ and its kharja entitled Dār aṭ ṭirāz fī ʿamal al muwas̲h̲s̲h̲aḥāt (دار الطراز في عمل الموشحات).'' The Syrian scholar published an edition of the work in 1949. Alan Jones published a modern edition of this work. An anthology of muwaššaḥāt entitled Uddat al Jalīs (عدة الجليس), attributed to a certain Ali ibn Bishri al Ighranati, is based on a manuscript taken from Morocco in 1948 by Georges Séraphin Colin (1893 1977). Alan Jones published an Arabic edition in 1992. Ibn Bassam wrote in (الذخيرة في محاسن أهل الجزيرة) that the kharja was the initial text around which the rest of the muwaššaḥ was composed. Ibn Khaldun also mentions the muwaššaḥ and its kharja in his Muqaddimah.

In 1948, the Hungarian linguist Samuel Miklos Stern published "Les Vers finaux en espagnol dans les muwaššaḥs hispano hebraïques" in the journal al Andalus, translated into English in 1974 as The Final Lines of Hebrew Muwashshaḥs from Spain. Stern's interpretation of kharjas in Hebrew texts made them accessible to Romanists and had a great impact on the Spanish establishment and scholars of Romance in the West. Gómez's 1965 book Jarchas Romances De La Serie Arabe En Su Marco presented a corpus of all known kharjas at the time; although it did not include annotation or scholarly apparatus, it became canonical. Solà Solé's Corpus de poesía mozárabe (Las Harjas andalusíes) offered a complete scholarly apparatus, variations taken from different manuscripts, thorough discussion, and thoughtful speculation. LP Harvey, Alan Jones, and James T. Monroe have also made influential contributions to the study of the kharjas.