Path to Discovery! Romanization & Scripts for Non-Latin/Arabic Materials - PowerPoint Presentation

Slide1

Path to Discovery!Romanization & Scripts for Non-Latin/Arabic MaterialsChallenges & Potential

Iman Dagher, UCLA

PCC Participants Meeting, January 26, 2020

ALA Midwinter Conference, Philadelphia, PA

Non-Latin Scripts in Libraries’ Data

Chinese 汉字 漢字

Arabic

ا

لعربية

Devanagari देवनागरी Bengali-Assamese বাংলা-অসমীয়াCyrillic Кириллица Kana かな カナ Hangul 한글 조선글 Telugu తెలుగు Tamil தமிழ் Gujarati ગુજરાતી

Kannada ಕನ್ನಡMalayalam മലയാളംThai ไทยGurmukhi ਗੁਰਮੁਖੀLao ລາວHebrew אלפבית Khmer ខ្មែរArmenian ՀայոցMongolian ᠮᠣᠩᠭᠣᠯBurmese မြန်မာ

2

ObjectivesRomanization: Challenges and PotentialRomanization: Challenges with Arabic materials Potential of

adding scripts Scripts within Name Authority File

Adding scripts retrospectively: UCLA project

3

PCC Practices

RDA instructs catalogers to transcribe data in the language and script found in the resource. The LC-PCC PS 1.4 is to apply the first alternative, i.e. to record the elements in a transliterated form

Follow the ALA-LC Romanization TablesUse of MARC Model A for Bibliographic

records



vernacular and transliteration The original-script fields are coded as 880 parallel fields in bibliographic recordsIn OCLC, parallel fields display as the same MARC tags as their linked Latin equivalentField 066 Use of MARC Model B for authority records  provides unlinked non-Latin script fieldsAdding scripts is optional but recommended4

Romanization in Practice: Advantages 

Romanized data is used by different tasks of the libraries 

Ability to integrate materials into library-wide processing and handling procedures: acquisition, shelf-listing, serials check-in, circulation, preservation, reference, interlibrary loan, etc.

Not all library staff are required to have a specialized language knowledge

Ability to find resources even if you don’t know the language

Some library systems do not support all non-Latin scripts5

Romanization in Practice: Challenges

Lack of consistent Romanization Romanization is not perfect: lack of consensus

Record duplicationsALA-LC Romanization criticism: ambiguous, complex, unrepresentative

Different Romanization tables: ALA-LC; ISO-based Romanization; German; etc.

Patron confusion and frustration in discovering and correctly citing library materials

Training Issues: library staff & patrons Cataloging challenges: searching and recording Need to maintain macrosRomanization issues with certain languages6

Arabic Script, Arabic Language: Some Facts

Arabic is one of the most widely used scripts in the world Around 660 million individuals use Arabic script to communicate in a number of languages, including: Urdu, Pashto, Arabic, Punjabi, Persian, Malaysian, and Kurdish

The Arabic language is a Semitic language with about 221 million speakers, and spoken in more than 34 countries

Over 30 different varieties of colloquial Arabic

Modern Standard Arabic is the universal language of the Arabic-speaking world, which is understood by all Arabic speakers

Arabic is written from right to left7

Romanization: Challenges with Arabic materials

Following certain rules in the ALA-LC Romanization tables for Arabic can be challenging

Requires special training and skills to master the ALA-LC Romanization table for the Arabic language (total of 26 rules)

Time-consuming with searching and locating the records (ISBN not always present or correct)

Different sources to consult; some sources such as Hans

Wehr suggest 2 Romanized forms for certain wordsRequires adding 246 (other forms of titles)Romanizing certain titles requires a familiarity with the culture Arabic language relies on Tashkil (or Tahrik), i.e. vocalization 8

Romanization: Issues with the Arabic Language

Tashkil is adding the different diacritics to Arabic letters in order to indicate vowels, or lack thereof, and gemination

Arabic texts are mostly written without tashkil in library resources. Fluent speakers are able to automatically fill in these diacritics themselves

Arabic is a highly inflected language. Romanization requires a good grammatical knowledge

©

Mamoun Sakkal 19979

Romanization: Issues with the Arabic Language

Romanizing foreign words:

Bibliography =ببليوجرافيا 

bibliyūjirāfiyā

bibliyu

̄jrāfiyā bibliyūjrāfyā = ببليوغرفيا  bibliyūghrafiyā bibliyūghrafyā bibliyūghirafiyā 10

Romanization: Issues with the Arabic LanguageThe position of

hamzah [

ء] on the

أ إ

will affect the beginning vowel  a, u, i 1. imām = leader  إمام amāma = in front  أمام 2. Africa: إفريقيا  Ifriqiyā OR أفريقيا

 Afriqiyā Erbil: أربيل  Arbīl OR إربيل  Irbīl [Irbīl in NAF] 11

Romanization: Issues with the Arabic LanguageGeographic names

Inconsistency America 

أميركا  Amīrikā; Amīrka

̄

أمريكا Amrīkā Authorized forms vs. commonly-used namesIn NAF: Rabat (Morocco) suggests the Romanizing as: Rabāṭ But most commonly known is: Ribāṭ الرِباط 12

Romanization: Issues with Arabic Personal Names

Sometimes it is difficult to predict how a name is pronounced with a lack of diacritics

حَسَن

Ḥasan name for male حُسْن Ḥusn name for femaleNames may be pronounced differently depending on the origin of the authorMunassá13

Romanization: Issues with Arabic Personal Names

Issue with normalization: searching for Salim can retrieve two different names

سليم [Salīm

]

سالم [Sālim]North African names have their special challenges with Romanization since pronunciation may be influenced by the French languageIssues with Latin names in Arabic script Shakespeare, Williamشكسبير، وليام More variants are recommended  Adding script variants can be the solutionSalimShakspīr, Walyam; Shakspīr, Wilyam; Shaykspīr,

Wilyam14

Romanization: Issues with Arabic Personal Names Authors who write in both Arabic & PersianDifferent romanization tables

Variant Romanized forms are necessaryAdding scripts would solve the issue

Example:Jawādi

̄

Āmuli

̄, ʻAbd Allāh (Arabic) Javādī Āmulī, ʻAbd Allāh (Persian)جوادي آملي، عبد الله15

Romanization: Issues with Arabic Personal Names: Duplicates

Jubrān, Thurayā, $d 1952- no2015104931

Jubrān, Thurayyā, $d 1952- n 2016006592

Jibrān

,

Thurayā, $d 1952- ʻImmīsh, Ibrāhīm Fatḥī no2011054221    ʻUmaysh, Ibrāhīm Fatḥi n 2011037693   جبران، ثريا، 1952 -عميش، ابراهيم فتحي16

Ḥamīd, ʻAbd al-Laṭīf ibn Muḥammad no 95050509 Ḥumayd, ʻAbd al-Laṭīf ibn Muḥammad n 93902168 Ḥumayyid, ʻAbd al-Laṭīf ibn Muḥammadحميد، عبد اللطيف بن محمد

010 no2012067640 $z no2006125278 $z nb2008017303100 1 Manīsī,

Aḥmad373 Markaz

al-Dirāsāt al-Siyāsīyah

wa-al-Istirātījīyah

$2

naf373 Markaz al-Imārāt lil-Dirāsāt wa-al-Buḥūth al-Istirātījīyah $2 naf375 male377 ara400 1 منيسي، أحمد 4001 Minīsī, Aḥmad400 1 Munīsī, Aḥmad400 1 Minnīsī, Aḥmad400 1 Munaysī, Aḥmad

400 1 Munayyisī, Aḥmad400 1 Menese, Ahmed667 Non-Latin script reference not evaluated.670 Niqābat al-Mihan al-Riyāḍīyah, 2004: $b t.p. (Aḥmad Manīsiأحمد منيسي = ̄)17

Value of Scripts for Arabic Materials Precision in discovery

More efficient cataloging practicesMore legible and understandable metadata for the patrons

Globalization

18

Adding Scripts: Considerations

Using Macros is very helpful, but additional review is needed Use of different scripts with different directionality in one field may affect the display

Not all scripts are available in authority file: e.g. Armenian

Display of scripts: problem with certain diacritics

19

UCLA Project: Adding Scripts to Legacy Data

Adding scripts in OCLC for monographs in Russian Cyrillic in bibliographic record fields held by UCLA Russian: generally a one-to-one transliteration

About 54 thousand records for Monographs (excluding mixed script records)

Using an in-house process, working with a library IT programmer, replacing the master records in OCLC

Fields: 245 ($

abcpn), 250, 26X ($abc), 490 $aSample records will be reviewed by librarians with language expertise The processed/replaced records in OCLC will be identifiable by a 588 field: “UCLA Machine-derived non-Latin script bibliographic record project” Future potential projects: Armenian; smaller number of records and easier to tackle/review20

Sources Consulted Agenbroad

, J. E. (2006) Romanization Is Not Enough, Cataloging & Classification Quarterly, 42:2, 21-34. retrieved from https://www.tandfonline.com/doi/pdf/10.1300/J104v42n02_03?needAccess=true

ALCTS Non-English Access Working Group on Romanization Report (2009, December 15).

http://www.ala.org/alcts/sites/ala.org.alcts/files/content/ianda/nonenglish/apd15a.pdf

Johnson

, Chr. B. (December 2012). An Introduction to the ALA-LC Romanization Tables, SCATNews, 38, retrieved from https://www.ifla.org/files/assets/cataloguing/scatn/scat-news-38.pdfOn Romanization (August 2019). Multilingual Library: A blog devoted to multilingual issues in the library catalog. Retrieved from https://elegantlexicon.com/libPCC Guidelines for Creating Bibliographic Records in Multiple Character Sets Report (2017, September). https://www.loc.gov/aba/pcc/bibco/documents/PCCNonLatinGuidelines.pdfVernon, E. (1996). Decision-making for automation : Hebrew and Arabic script materials in the automated library. Retrieved from https://www.ideals.illinois.edu/handle/2142/3879The World's Most Popular Writing Scripts (October 23, 2019). Retrieved from https://www.worldatlas.com/articles/the-world-s-most-popular-writing-scripts.html 21

شُكْراً

Shukran

Thank you

22