SEAlang Library Malay Dictionary
About the SEAlang Library Malay Resources 
SEAlang's Malay dictionary is primarily based on the works of Sir Richard James Wilkinson. 
-- A Malay-English Dictionary (romanised), 1932, Mytilene, Greece: Salavopoulos and Kinderlis. 1,288 pages, 26,000 plus heads, nearly 12,000 references, 21,500 etymological notes, 9,000 plus citations of authorities, 4,000 formal species references. 
-- An Abridged Malay-English Dictionary (romanised), Third Edition, 1926, Singapore: Kelly & Walsh. 2,500 etymological notes, 11,000 heads, 3,500 derivatives, 4,000 compounds.
A less formal contemporary work is included to help ensure that modern spellings and abbreviations can be found:
-- Bhanhot's Malay-English Dictionary is a pioneering (1996) effort by Dr. Diljeet Kumar Bhanot.  It contains roughly 6,000 heads and 3,500 subheads.  A former chairman of the Malaysian Medical Association, Selangor, Dr. Bhanot is in private practice; see
    Like other Austronesian languages, Malay consists mostly of lemmas (that is, dictionary headwords) that have been prefixed, suffixed, or reduplicated (lemmas, unlike roots, can generally stand alone).  Traditionally, all forms are listed under the lemma, which can make it difficult to find unfamiliar words.  We use a variety of techniques to automatically find useful information, including this rule-based lemmatiser developed by Timothy Baldwin and Su'ad Awab.
A note on transcription and tagging of the "Unabridged"
Normally the SEAlang Lab will individually tag all usable information, including compounds, references, loan sources, etc.  Inconsistencies and idiosyncracies in the original publication (for example, abbreviations are often context-dependent, and thousands of Latin terms and loan sources are printed with spaces between each letter) made this impossible for the Wilkinson Unabridged.
    We have put our efforts into tagging the parts of the text that seem most useful to the scholarly community, and provide entry-by-entry links to the original text image to allow immediate checking of the transcription (please bring any errors to our attention).  If anybody wishes to pursue developement of an edited, critical, or updated edition of Wilkinson, we would be delighted to assist. 
Wilkinson's Dictionaries
Encyclopediaic in scope, Wilkinson's 1932 edition is the culmination of decades of work on Malay, which included a first 'unabridged' (of about half its size) published in 1901-1902. 
    Wilkinson's orthography presents a slight problem.  For much of the 20th century Wilkinson established the standard for romanization, enabling the transition from the previously universal Arabic-based Jawi script.  A series of spelling reforms, culminating in the 1972 harmonization of Malay and Indonesian, has left Wilkinson slightly out of step with modern practice.  As a rule, Wilkinson's ch = modern c, his ĕ (which reflected ə, and which you can enter with E) = e, and - (for hyphenation of affixes) is not used. 
    We provide a checkbox (under Match) that allows for approximate matching of these.  However, other modern alternations (e.g. o/u and e/i) are not handled automatically at this time; please contact us if you'd like to help gloss Wilkinson's spelling. 
Please note the alternatives for display on the left:
-- show self only shows the headword or derived form that matched (try beranak).
-- show self/parent shows the lemma of the (derived) form as well (try beranak).
-- show family shows the lemma and all derived forms (try anak).

Searches may:
-- match a single character with ?, and zero or more with *. Thus, ca?ang matches cabang, cadang, canang, while ca*ang matches cadangan, cawangan, campur tangan and many others.
-- require matches for both, or either, Indonesian text (of the headword), or English text (in the definition).
-- be limited to a particular etymology, usage, type, or subject.
Predictive completion    Dictionary heads will appear as you type.  Short lists will include multi-word entries and examples (compare ban to band).
Lemmatization and "smart" search    To help locate phrases and derivatives, search will automatically:
-- search examples for an instance of the word if it isn't found in a head (try anakan).
-- next, lemmatize the word, and search for the root.
-- In case of a multi-word phrase, we automatically search examples, then search for the phrase as a list of separate words, then search for the list of roots.
The English search term can be expanded (default) to include inflected forms (a search for sing matches sings, singing, sang, sung as well). Approximate search for Javanese means that we ignore diacritics, glottal marks (the ' ), and hyphens in compounds.