In this post, I’d like to provide an overview of the ways Islamic names encountered in AML/KYC contexts can differ from familiar Anglo/European names. I will emphasize those aspects that cause complications both for automated processing (searching and matching) and for human interpretations. In several previous posts, I have covered Hispanic names, Korean names, and Russian names. For a more complete discussion of the issues with Hispanic, Islamic, Korean, and Russian names, download our white paper on Improved OFAC Name Screening.]
Islamic vs. Arabic
First, it’s important to remember that Islamic is not the same as Arabic, any more than Anglican is the same as English. In fact, more members of the Islamic faith live outside the Arabic world than in it. And, as with Judaeo-Christian religions, the Islamic faith has a pervasive impact on the naming practices followed by its believers, whether they are Arab or not.
Seven Things That Make Islamic Names Difficult
What exactly is it about Islamic names that makes them so opaque and challenging to the vast majority of Westerners, especially in an AML/KYC context? Here are seven basic factors that make it hard for most westerners to understand Islamic names.
Alphabetic vs. Abjad Writing Systems
Arabic script is written using an “abjad” system, meaning consonant sounds are emphasized, and vowels are generally left for the reader to infer. In everyday use, Islamic names represented in Arabic script would rarely appear in “vowelized” form, so there is much more variation in the way a name from an Islamic country might appear, once it has been converted into the Roman alphabet system, where both consonants and vowels are included.
“Romanization” is the name for a process that converts a written item from its native writing system into the Roman alphabet (A-Z). It is a specific form of transliteration, the process of moving linguistic data between and among various representation systems. Much of the name-data that fall within the scope of modern automated banking and financial systems is created originally in the Roman alphabet, but some originates in languages like Arabic. Although there are well-defined ways to Romanize data written in Arabic, these standards tend to apply more loosely and inconsistently to personal names, meaning that the same name can be represented in a surprisingly wide variety of ways.
Anglophone vs. Francophone
As a corollary of the previous point, the two most commonly encountered forms of Romanization apparent in Islamic names appearing in compliance data are the ones associated with the European cultures that have arguably had the biggest impact on events in the Middle East: England and France. These two styles of Romanization have significant differences when applied to the same Islamic name, as represented in the Arabic writing system.
Here are just a few of the most evident ways that these two widely used Romanization systems can cause the same Islamic name to appear in very different spelled forms:
- French OU – The French-style OU is complicated in Islamic names from the Maghreb (Western) countries of North Africa, such as Morocco, Algeria and Tunisia. Sometimes OU is found where the vowel U would occur in an Anglo-style Romanized name (OUSMANE-USMAN; OUSAMA-USAMA), but sometimes it corresponds instead to W, a consonant (OUALIDE-WALID; MEROUANE-MARWAN), which can cause serious problems for many character-based name-matching techniques, including Soundex.
- English J/JJ – The “zh” sound generally represented by the letter J in English name Romanizations (HAJ, HAJJ) generally appears as GE in the equivalent French form (HAGE-HAJ) , but also may appear as DJ (DJAMEL-JAMAL).
- English S – Orthographic S in English versions can appear instead as C in the French equivalent form (YOUCEF-YUSSEF)
- English SH – SH in Anglo-style Romanized forms will generally be found instead as CH in the equivalent French form (RACHID-RASHID; CHEIKH-SHEIKH; HICHAM-HISHAM).
- English CH – Where English Romanized versions of Islamic names have CH, the French equivalents typically have TCH instead (TCHALABI-CHALABI).
Additional spelling differences in Maghrebi Islamic names can also be introduced by dialectal variations in the way Arabic is spoken in that area.
The 99 Names of Allah and the “Servant-of” Name-Template
The 99 Names of Allah provide a popular and frequent basis for many Islamic personal names. For example the name Romanized as ABD AL RAHMAN is identified as “Servant of the Compassionate One” and ABDAL AZIZ is “Servant of the Almighty One,” to name just a few of the more popular attributes.
Articles and White-space
Still more inconsistencies can arise from the way in which Islamic names render the Arabic definite article (ال in Arabic, like THE in English) and the way that white-space (blanks) appear between and among the various parts of a name. This highly ubiquitous component of Islamic names may appear as AL, EL or UL. But the final letter, L, can also be converted to R, S or D in written forms, because it assimilates (is influenced by) the first sound of the following name-part (ABDEL RAHMAN – ABDERRAHMAN; ABDEL SALAAM-ABDESSALAAM; NUR EL DIN-NUREDDINE).
Patrilinear Elements vs. Surnames
Family names and surnames exist in the Islamic naming system, but a surname that is used as part of the public identity may not be based on a family name. Instead, references to various male forebears may be used, thereby taking on the same function as a surname does in Western society. Confusion by westerners of the nasab (patronymic, patrilinear) name-elements and the nisbah (familial, tribal) name-elements means, as a practical matter, that individuals (especially males) with Islamic names have some degree of life-long discretionary control over the part of their names that will ultimately be regarded as the surname, especially in its Romanized form.
Son-of vs. Father-Of in Male Names
Islamic names can be influenced not only by ancestry, but also by progeny. That is, Islamic names can not only refer to the males who preceded the bearer, but also to a child of the bearer. This portion of the name is called a kunya, and it may be present in both male and female personal names. For name-matching purposes, the kunya complicates the process of tracking the same individual across time, as he or she may proceed from single life to parenthood, and then have multiple children, since the kunya adopted by the bearer may change at each time his or her parenthood status changes. This is especially impactful in Romanized name-forms where the kunya may (like the nasab) come to be taken as the surname.
The points covered briefly in this post are by no means an exhaustive treatment of this very complex topic. Instead, they are intended to convey a sense of the many reasons, large and small, why Islamic names pose an especially great challenge for AML/KYC professionals whose ability to recognize or infer relationships between names is a key job qualification.
For a more extensive analysis of the problems inherent in working with Romanized Islamic names, please see the White Paper on this topic.