Beruflich Dokumente
Kultur Dokumente
i
COLLABORATORS
TITLE :
Urdu Morphology
REVISION HISTORY
ii
Contents
1 Introduction 1
1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Dual-use Grammars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Audience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.4 More on Uses of this Grammar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4.1 Grammar as a Basis for Computational Tools . . . . . . . . . . . . . . . . . . 3
1.4.1.1 Uses of a Parser and Generator . . . . . . . . . . . . . . . . . . . . . 4
1.4.1.2 Building a Parser and Generator . . . . . . . . . . . . . . . . . . . . 5
1.4.2 Grammatical Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.5 Scope of this Grammar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.6 Looking to the Future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.6.1 Grammar Help . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.6.2 Spell Correction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.6.3 Grammar Adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.6.3.1 Model Grammars . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.6.3.2 Automated Grammar Adaptation . . . . . . . . . . . . . . . . . . . . 12
1.7 Format of this Grammar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
iii
2.2.2.1 Vowel Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.2.2.2 Short Vowels with Diacritics . . . . . . . . . . . . . . . . . . . . . . 18
2.2.2.3 Long Front Vowels with یyē . . . . . . . . . . . . . . . . . . . . . . 19
2.2.2.4 Long Back Vowels with وvāū . . . . . . . . . . . . . . . . . . . . . . 19
2.2.2.5 Other Ways of Writing Vowels . . . . . . . . . . . . . . . . . . . . . 20
2.2.2.6 Vowel Mutations Before /h/ and ‘ain . . . . . . . . . . . . . . . . . 20
2.2.2.7 Nasal Vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2.3 Orthography of Consonants . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2.3.1 Transcription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2.3.2 Consonantal Diacritics . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.2.4 Spelling Issues in Urdu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.2.4.1 Vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.2.4.2 Consonants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.2.5 Table of Characters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3 List of Abbreviations and Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3 Nouns 37
3.1 Inflection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.1.1 Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.1.1.1 Masculine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.1.1.1.1 Masculine Marked (Class I) . . . . . . . . . . . . . . . . . . 39
3.1.1.1.2 Masculine Unmarked (Class II) . . . . . . . . . . . . . . . . 39
3.1.1.2 Feminine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.1.1.2.1 Feminine Marked (Class III) . . . . . . . . . . . . . . . . . 39
3.1.1.2.2 Feminine Unmarked (Class IV) . . . . . . . . . . . . . . . . 40
3.1.1.3 Some Clues to Gender of Unmarked Nouns . . . . . . . . . . . . . . 40
3.1.1.3.1 Typical Feminine Suffixes and Patterns . . . . . . . . . . . 40
3.1.1.3.2 Typical Masculine Suffixes and Patterns . . . . . . . . . . . 41
3.1.2 Inflectional Affixes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.1.2.1 Masculine Nouns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.1.2.2 Feminine Nouns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.1.3 Formal Grammar of Nouns . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.1.4 Noun Paradigms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.1.4.1 Masculine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
iv
3.1.4.1.1 Marked . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.1.4.1.2 Unmarked . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.1.4.2 Feminine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
3.1.4.2.1 Marked . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
3.1.4.2.2 Unmarked . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
3.1.5 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
3.1.6 About Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.1.6.1 Vocative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.1.6.2 Direct . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.1.6.3 Oblique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.2 Nominal Derivational Suffixes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.2.1 وٹ/-vat ̣/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.2.2 ؤ/-ō/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.2.2.1 Masculine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.2.2.2 Feminine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.2.3 ی/-ī/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.2.3.1 ی/-ī/ added to stems of simple causative verbs . . . . . . . . . . . 56
3.2.3.2 ی/-ī/ added to stems of simple or double causative verbs . . . . . 56
3.3.2 Arabic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.3.2.1 Nouns from Arabic . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.3.2.1.1 Place Nouns . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.3.2.1.2 Instrument Nouns . . . . . . . . . . . . . . . . . . . . . . . 63
v
3.3.2.1.3 Abstract Nouns . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.3.2.1.4 Arabic Definite Articles in Urdu . . . . . . . . . . . . . . . 63
3.3.2.1.5 Noun Plurals . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.3.2.1.5.1 Dual Forms . . . . . . . . . . . . . . . . . . . . . . . 64
3.3.2.1.5.2 Regular (“Sound”) Plurals . . . . . . . . . . . . . . . 64
3.3.2.1.5.3 Irregular (“Broken”) Plurals . . . . . . . . . . . . . . 65
3.3.2.2 Arabic Prefixes in Urdu . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.3.2.3 و/-o-/ and Arabic and Persian conjunctive particle . . . . . . . . . . 67
4 Pronouns 68
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.2 Personal Pronouns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.2.1 Regular Forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.2.2 Alternate Forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.2.3 Possessive Personal Pronouns . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.3 Reflexive Pronouns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.3.1 د/xud/ and آپ/āp/: X-self . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.3.2 6 ا/apnā/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.3.3 د/xud/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
4.3.4 آپ ا/apnē āp/ and د9 د/xud baxud/ . . . . . . . . . . . . . . . . . . . . . 74
5 Postpositions 82
5.1 Grammatical Postpositions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
vi
5.1.1.1.2 Types of Possession . . . . . . . . . . . . . . . . . . . . . . 84
5.1.1.2 Compound Postpositions Using 5 /kā/ . . . . . . . . . . . . . . . . . 85
5.1.1.2.1 5 /kā/ + oblique noun + postposition . . . . . . . . . . . 85
5.1.1.2.2 5 /kā/ + oblique noun . . . . . . . . . . . . . . . . . . . . 87
5.1.1.2.3 5 /kā/ + oblique adjective . . . . . . . . . . . . . . . . . . 88
5.1.1.2.4 5 /kā/ + adverb . . . . . . . . . . . . . . . . . . . . . . . . 89
5.1.1.2.5 ( 5 /kā/) + Perso-Arabic preposition + oblique noun . . . 90
5.1.1.3 With Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.1.2 @ /kō/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.1.2.1 Indirect Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.1.2.2 Direct Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.1.2.3 Subjects (Indirect Constructions) . . . . . . . . . . . . . . . . . . . . 93
5.1.2.4 Possession of Intangible Objects . . . . . . . . . . . . . . . . . . . . 94
5.1.3 /nē/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5.2 Spatial-Temporal Postpositions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
5.2.1 /sē/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
5.2.1.1 Ablative Use (Place and Time) . . . . . . . . . . . . . . . . . . . . . 96
5.2.1.2 Instrumental/Agentive Use . . . . . . . . . . . . . . . . . . . . . . . 96
5.2.1.3 Comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
5.2.1.4 Adverbial Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.2.1.5 With Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.2.1.6 Postpositional Sequences . . . . . . . . . . . . . . . . . . . . . . . . 98
5.2.2 /tak/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.2.3 ب/mēṁ / . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.2.3.1 Locative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.2.3.2 Change of State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.2.3.3 Cost in Time/Money . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.2.3.4 After Infinitives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.2.4 ب/par/ ~ بہ/peh/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.2.4.1 Locative Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.2.4.2 Motion Towards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.2.4.3 After Infinitives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.2.4.4 With Objects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
vii
5.3 Verb + Postposition Collocations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.4 Prepositions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
5.4.1 Persian Prepositions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
5.4.2 Arabic Prepositions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
6 Adjectives 105
6.1 Inflection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.1.1 Marked (Inflected) Adjectives . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.1.1.1 Masculine and Feminine . . . . . . . . . . . . . . . . . . . . . . . . 105
6.1.1.2 Direction and Numbers . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.1.1.3 Agreement with Multiple Nouns . . . . . . . . . . . . . . . . . . . . 109
6.1.2 Unmarked (Uninflected) Adjectives . . . . . . . . . . . . . . . . . . . . . . . 110
6.1.2.1 Persian Past Participles in cʰot ̣ī hē . . . . . . . . . . . . . . . . . . . 110
6.1.2.2 Arabic Adjectives in ی/-ī/ or ی/ ا/-ā/ . . . . . . . . . . . . . . . . 111
6.1.2.3 Denominal Adjectives . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.1.2.3.1 In ی/-ī/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.1.2.3.2 In ابہ/-ānah/ . . . . . . . . . . . . . . . . . . . . . . . . . . 112
viii
6.5.2 Phrasal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
6.5.2.1 With /sē/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
6.5.2.1.1 Comparatives . . . . . . . . . . . . . . . . . . . . . . . . . . 118
6.5.2.1.2 Superlatives . . . . . . . . . . . . . . . . . . . . . . . . . . 118
6.5.2.2 With ز دہ/zyāda/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
6.6 Phrasal Adjectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
6.6.1 With /sā/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
6.6.1.1 Noun/Oblique Pronoun + /sā/ . . . . . . . . . . . . . . . . . . . 118
7 Miscellaneous 121
7.1 Adverbs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
7.1.1 Demonstratives and Interrogative and Relative Adverbs . . . . . . . . . . . . 121
7.1.1.1 Interrogatives vs Relatives . . . . . . . . . . . . . . . . . . . . . . . 123
ix
7.2.1.3.3 With Relatives . . . . . . . . . . . . . . . . . . . . . . . . . 131
7.2.2 Conjunctions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
7.2.2.1 Coordinating Conjunctions . . . . . . . . . . . . . . . . . . . . . . . 132
7.2.2.1.1 اور/ɔr/ and . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
7.2.2.1.2 /yā/or . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
x
7.2.5.2.2 Free Interjections . . . . . . . . . . . . . . . . . . . . . . . 146
7.3 Repetition and Reduplication in Urdu . . . . . . . . . . . . . . . . . . . . . . . . . . 146
7.3.1 Repetition and Reduplication of Nouns . . . . . . . . . . . . . . . . . . . . . 146
7.3.1.1 Repetition of Nouns . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
7.3.1.2 Reduplication of Nouns (Echo Words) . . . . . . . . . . . . . . . . . 146
7.3.2 Repetition of Pronouns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
7.3.3 Repetition and Reduplication of Adjectives . . . . . . . . . . . . . . . . . . . 148
7.3.3.1 Repetition of Adjectives . . . . . . . . . . . . . . . . . . . . . . . . . 148
7.3.3.2 Reduplication of Adjectives . . . . . . . . . . . . . . . . . . . . . . . 148
7.3.4 Repetition of Adverbs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
7.3.5 Repetition of Verbal Forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
7.3.5.1 Repetition of Conjunctive Participles . . . . . . . . . . . . . . . . . 149
7.3.5.2 Repetition of Imperfective Participles . . . . . . . . . . . . . . . . . 149
7.3.5.3 Repetition of Perfective Participles . . . . . . . . . . . . . . . . . . . 151
8 Verbs 152
8.1 The Four Basic Verb Forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
8.1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
8.1.2 The Stem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
8.1.3 The Imperfective Participle . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
8.1.4 The Perfective Participle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
8.1.5 The Infinitive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
8.2 The Auxiliary Verb /hōnā/̄ to be . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
8.2.1 Present Tense . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
8.2.2 Simple Past Tense . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
8.2.3 Subjunctive Mood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
8.2.4 Future Tense . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
8.2.5 Formal Grammar of /hōnā/̄ to be . . . . . . . . . . . . . . . . . . . . . . . . . 160
8.3 Verb Constructions Using the Stem . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
8.3.1 Imperatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
8.3.1.1 Intimate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
8.3.1.2 Familiar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
8.3.1.3 Formal/Polite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
8.3.1.3.1 Regular Forms . . . . . . . . . . . . . . . . . . . . . . . . . 166
xi
8.3.1.3.2 Irregular Forms . . . . . . . . . . . . . . . . . . . . . . . . 166
8.3.2 Subjunctive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
8.3.2.1 Forming the Subjunctive . . . . . . . . . . . . . . . . . . . . . . . . 167
8.3.2.1.1 Regular Subjunctive Conjugation . . . . . . . . . . . . . . . 167
8.3.2.1.2 Irregular Subjunctives . . . . . . . . . . . . . . . . . . . . . 167
8.3.2.2 Uses of the Subjunctive . . . . . . . . . . . . . . . . . . . . . . . . . 167
8.3.2.2.1 Subjunctive in Main Clauses . . . . . . . . . . . . . . . . . 169
8.3.2.2.2 Subjunctive in Subordinate Clauses . . . . . . . . . . . . . 171
8.3.2.2.2.1 A Note on Conditional Sentences . . . . . . . . . . . 174
8.3.3 Future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
8.3.3.1 Forming the Future . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
8.3.3.2 Uses of the Future . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
8.3.3.2.1 In Main Clauses . . . . . . . . . . . . . . . . . . . . . . . . 175
8.3.3.2.2 In Conditional Sentences . . . . . . . . . . . . . . . . . . . 175
8.3.4 Conjunctive Participles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
8.3.5 Progressive Tenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
8.3.5.1 Progressive Present . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
8.3.5.2 Progressive Past . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
8.3.5.3 Progressive Subjunctive . . . . . . . . . . . . . . . . . . . . . . . . . 180
8.3.5.4 Progressive Presumptive . . . . . . . . . . . . . . . . . . . . . . . . 182
8.3.5.5 Progressive Irrealis . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
8.4 Verb Constructions Using the Imperfective Participle . . . . . . . . . . . . . . . . . . 184
8.4.1 Bare Participle Forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
8.4.1.1 Irrealis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
8.4.1.2 Ellipsis in the Narrative Imperfective . . . . . . . . . . . . . . . . . 185
8.4.2 Imperfect Tenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
8.4.2.1 Present Imperfect . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
8.4.2.2 Past Imperfect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
8.4.2.2.1 Narrative Imperfective . . . . . . . . . . . . . . . . . . . . 187
8.4.2.3 Subjunctive Imperfect . . . . . . . . . . . . . . . . . . . . . . . . . . 187
8.4.2.4 Presumptive Imperfect . . . . . . . . . . . . . . . . . . . . . . . . . 189
8.4.2.5 Irrealis Imperfect . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
8.4.2.6 As Adjectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
xii
8.5 Verb Constructions Using the Perfective Participle . . . . . . . . . . . . . . . . . . . 192
8.5.1 Miscellaneous Functions of the Perfective Participle . . . . . . . . . . . . . . 192
8.5.1.1 Simple Perfective . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
8.5.1.2 Perfective Participle + /jānā/ . . . . . . . . . . . . . . . . . . . . 193
8.5.1.2.1 Passive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
8.5.1.2.2 Incapacity . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
xiii
8.7.4 ر/rahnā/: Continuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
8.7.5 ر/rahnā/: Emphatic Continuation . . . . . . . . . . . . . . . . . . . . . . . 213
8.7.6 /jānā/: Progression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
Index 254
xiv
List of Tables
xv
4.2 Personal Pronouns: Alternate Forms (Special Dative) . . . . . . . . . . . . . . . . . . 72
4.3 Demonstrative Pronouns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.4 Relative Pronouns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.5 Interrogative Pronouns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.1 5 /kā/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
xvi
8.20 Imperfect Subjunctive Paradigm of : /karnā/ to do . . . . . . . . . . . . . . . . . . 188
xvii
Introduction
Chapter 1
Introduction
1.1 Overview
This book describes the grammar of Urdu, with emphasis on its morphology; the description is
supplemented by a formal grammar, written in XML, which may be used to implement a morpho-
logical parser. It is not intended as a pedagogical grammar.
This chapter describes how this grammar is structured, and for what purposes it is intended;
included is a brief description of the use of the formal grammar, and the process for converting it
into a morphological parser. The remaining chapters constitute the grammar itself, with a focus
on the morphology, or word structure, of Urdu.
This grammar may be used in several ways:
• As a grammar which is easily converted into computational tools, such as morphological parsers;
These intended uses are described in more detail in Section 1.2 of this chapter.
The grammar is structured into two separate but parallel and intertwined grammars: one is a
traditional description of the morphology of Urdu, in a form that a researcher or engineer with a
minimum of training in formal linguistics would understand. The other grammar component is a
formal grammar of Urdu morphology, suitable for automated extraction and conversion into the
form a computer program would use. The relationship between these two grammars is described
in Section 1.2 of this chapter.
1
Introduction
This grammar of Urdu is tailored to its intended use as a platform- and application-independent
way of representing the grammar, and particularly the morphology, of Urdu in a way that will
make it straightforwardly usable as the basis for computational tools. (We describe the method
for converting this grammar into a form usable by computational tools later in this chapter.)
This report constitutes a “Dual-Use” grammar, meaning that the grammar is represented twice,
for two distinct but related purposes: once in descriptive terms, using English, and once in formal
terms, using an XML-based model.1
The two descriptions are woven together, the intention being that the strengths of each will support
the weaknesses of the other. In particular, a grammar written in English tends to be ambiguous,
whereas the formal grammar should be unambiguous. Our expectation is that where the English
descriptions are (unintentionally) ambiguous, referring to the formal grammar will disambiguate
our intended meaning. At the same time, a weakness of formal grammars (and particularly formal
grammars that are computer-readable, as is ours) is that they tend to be difficult for humans to
comprehend. Again, our intention is that the meaning of the formal grammar will be clarified by
the descriptive grammar.
1.3 Audience
The dual-use grammar presented here is in a format that should be useful to knowledge providers
ten or one hundred years from now, should they need to re-implement the grammar for Urdu on
a new platform or computer environment.
The primary intended audience of this report is a computational linguist, that is, a person who has
skills in both the science of linguistics and the art of creating computational tools for linguistic
analysis (including some familiarity with XML). Since the grammar is especially concerned with
the morphology of Urdu, and in particular with supporting the creation of morphological analysis
tools, the application of this grammar assumes that the computational linguist is knowledgeable
about technology for morphological analysis, and in basic linguistic terminology for morphology
(and to a lesser extent, the terminology for phonology). For such an implementation-oriented
knowledge provider, both parts of the dual-use grammar will be of interest, although the descrip-
tions of the actual usage (as opposed to form) of various affixes or constructions will probably be
of lesser concern. However, the examples in the paradigm tables and the examples of usage can
be used for testing parser implementations.
On the other hand, a linguist who wants to understand Urdu morphology will likely find the de-
scriptions of usage to be essential. These linguists (whether government or university researchers
in linguistics) who wish to learn about the grammar (and particularly the morphology) of Urdu,
form a secondary audience for this report. Since such a linguist is primarily concerned with un-
derstanding how Urdu morphology works, and with the meaning of affixes or constructions, the
formal grammar may be of less interest, except insofar as it may be necessary to refer to the formal
1
XML stands for ‘Extensible Markup Language,’ and is a computer representation of text in which the function of
pieces of text is indicated by “tags.” Our reasons for choosing this representation for the grammar are described in
an earlier technical report, TTO 1308 Technical Report M.5 “Standards for Lexical and Morphological Interchange,”
available from the University of Maryland Center for Advanced Study of Language (CASL).
2
Introduction
grammar to disambiguate the descriptive grammar. We discuss the direct use of the grammars by
a language analyst in Section 1.6.1.
Another audience we have tried to keep in mind is the linguist who is charged with describing the
morphology of another language, particularly of a language related to Urdu. Such a person may
wish to adopt the model given here to this other language. Depending on the purpose, either the
descriptive grammar or the formal grammar, or both, may be of interest in this context.2
The writing system normally used to write the Urdu language is the Urdu script (described be-
low). The differences between the Urdu script and more familiar writing systems may occasion
difficulties for the reader. We have therefore supplemented the examples written in Urdu script
with a transcription into a Latin-based script, as discussed in Chapter 2. For some purposes (such
as transcription of spoken Urdu), the Latin script may be sufficient, and readers who wish to can
safely ignore the Urdu script.
In this section we describe in more detail the potential uses we see for this grammar.
This report was produced in the Extensible Markup Language (XML), which provides a mecha-
nism for describing the structure (as opposed to the display format) of documents. Specifically,
this report is structured as a DocBook XML document; DocBook is a form of XML that has been de-
veloped for book- and article-like documents, particularly technical documents like this one. The
DocBook formalism, and the modifications to that formalism that we have used in this project, is
described in more detail in Section 1.7.
For some purposes (such as converting the grammar into a form suitable for use with computa-
tional tools, as described in the next sub-section), the native XML is the appropriate format. But
for other purposes, such as reading the report as a grammatical description, it is convenient to
format it for viewing by converting the XML tags formatting to appear on-screen or on a printed
page. This can easily be done by a variety of means, since the DocBook XML format is a widely
used format, and many tools are available for conversion.
3
Introduction
A morphological parser is a computational tool which analyzes words into their meaningful parts.
For example, a parser that uses a grammar and lexicon (dictionary) of English could analyze the
word ‘cats’ into the noun ‘cat’ and the plural suffix ‘-s,’ or the irregular verb ‘kept’ into the verb
‘keep’ and a past tense suffix. While this is simple to do for English, because English morphology
is quite simple,4 it can be much more complex for other languages.
In theory, a grammar intended to be used by a computer would not need a descriptive component,
written in English; a formal grammar, written in some unambiguous format, would suffice. In
fact, many such formal grammars have been written for a variety of languages already. We have
opted to blend a formal grammar with a descriptive grammar, using the technology of Literate
Programming (described in Section 1.7), thereby making this grammar understandable, as well as
unambiguous to make it portable to future computing environments.
There are many uses to which a morphological parser and generator can be put. The primary ap-
plication envisioned in this project is making dictionary lookup easier by automating the process
of converting an arbitrary inflected form of a word into the form under which it is listed in the
dictionary (the headword, or citation form). For some words, this is a trivial operation: simply
stripping off a suffix string. But for other words it can be considerably more complicated, involv-
ing, for example, changes to the pronunciation and spelling of the stem (such as the changes in
the English pairs keep-kept and break-broke).
A morphological parser can also serve as part of several fully automated applications, including a
gisting or machine translation system.5 While the morphological parsing step is unnecessary for
machine translation of some languages, because their morphology is so trivial, most languages
have at least some morphology that needs to be separated off and understood, both in order to
translate the meaning of the affixes, and in order to find a dictionary citation form so that the
word can be looked up in the translation dictionary, or otherwise converted into English.
Morphological parsing may also be useful in cross-language information retrieval, in which a user
uses a query posed in one language to search through a set of documents written in a foreign
language. For example, a user might know only English, but wish to search for documents written
in Urdu (presumably so that relevant documents can be selected for translation.)
Finally, a morphological generator can be used to produce all the inflected forms of a particular
word. For example, given a suitable grammar of English morphology, a morphological generator
could produce from the English word ‘break’ the forms ‘breaks,’ ‘breaking,’ ‘broke,’ and ‘broken.’ A
4
Specifically, the inflectional morphology of English—the part of English morphology that deals with things like the
plural of nouns and the past tense or -ing form of verbs—is simple. There is another part of English morphology, called
derivational morphology, which deals with affixes that, for example, change verbs into nouns (such as the derivation
of the noun ‘destruction’ from the verb ‘destroy’). English derivational morphology is not so simple.
5
A machine translation system is a computer program that automatically translates texts from one language to
another. While the technology is improving, a machine translation is not as good as what can be produced by a good
human translator. A gisting program is one step less than machine translation, in the sense that its translation is not
as good as what a “real” machine translation program could produce. Nevertheless, gisters have an advantage over
machine translation programs in that they are much easier to produce, in terms of human and computer resources,
and in terms of time. Gisters can therefore find use in document triage, an application in which a human who does
not know the language being translated is able to sort through a large number of gisted documents to select those for
which it appears useful to expend more energy on, e.g. by getting a translation from a human translator.
4
Introduction
chart of the inflected forms is called a “paradigm,” and is frequently used in teaching the grammar
of foreign languages.6
Using this grammar for computational implementation requires several steps.7 First, we extract
the formal grammar from the grammatical description. This operation has been programmed as
a simple XSLT (Extensible Stylesheet Language Transformation), which operates on the complete
grammar to extract the formal grammar, and reordered to conform to a set of XML schemas. These
schemas model common linguistic structures, such as ‘affix,’ ‘phonological rule,’ and ‘allomorph,’
in XML, and were created for this project based on previous work modeling linguistic structures.
The output of this process is the formal grammar in XML format.8
Second, this extracted XML grammar is translated into the programming language of the cho-
sen morphological parsing engine. This conversion can be done by any program which deals with
XML and allows conversion into other formats. We have implemented our converter in the Python
programming language. We chose Python because it allows the use of an object-oriented program-
ming approach, in which each linguistic structure expressed in the XML grammar corresponds to
a ‘class.’ For example, there are classes in the converter (and in the XML schemas) corresponding
to ‘affix,’ ‘phonological rule,’ and ‘allomorph.’ The result is that approximately half the converter
program—that part of the program devoted to modeling the linguistic structures—is reusable re-
gardless of the morphological parsing engine being targeted. Furthermore, the entire converter
program is generic in terms of the language being described: that is, the same converter will work
for a grammar of Bengali, Urdu, or any other language for which an XML grammar conforming to
the schemas has been written.
The part of the converter that is specific to a particular morphological parsing engine is the part
that rewrites the grammar into the programming language of that parsing engine. Our converter
currently targets the Stuttgart Finite State Transducer tools,9 10 Targeting a different parsing en-
gine would require rewriting this half of the converter, so that the grammatical structures defined
in the XML-based grammars could be used with the new parsing engine.11
6
An example of this is the ‘501 X Verbs’ books, where X is one of Spanish, French, German, Italian etc. While
learners frequently find such books useful, they are only available for major languages; nor are there corresponding
books giving the declensions of nouns and adjectives.
7
The conversion process, together with the programs which perform it, is described in a separate CASL TTO 1308
deliverable.
8
An example of a piece of this formal grammar structure in XML form is given in Sec-
tion 1.4.2. The previous modeling work from which our XML schemas were adapted is described at
http://fieldworks.sil.org/ModelDoc/ModelDocumentation.chm.
9
The Stuttgart Finite State Transducer (SFST) is an open source program, available from http://www.ims.uni-
stuttgart.de/projekte/gramotron/SOFTWARE/SFST.html; it supports the kinds of constructions needed for Urdu and
many other languages. Finite state transducers combine parsing and generation capabilities,
10
While the converter accounts for virtually all the linguistic constructions needed for Urdu, there are some linguistic
constructs in other languages, such as infixes and reduplication, which are not yet handled.
11
Targeting a parsing engine which had similar capabilities to the Stuttgart Finite State Transducer, such as the Xerox
Finite State Transducer, would probably mean rewriting less than half of the converter code, since the programming
languages are similar enough that there would be significant re-use of the half of the converter that is specific to the
Stuttgart parser.
5
Introduction
The final step is to use the parsing engine to compile the converted grammar, together with an
electronic dictionary of the language.12
In summary, the XML-based grammars serve as a stable way to define the morphological analysis
of natural languages such as Bengali or Urdu, so that the grammars can be used by different
parsing engines. The converter can be used for any language for which the morphology has been
described using the formal grammar. When it is desired to build parsers using a different parsing
engine, only a part of the converter needs to be changed; the grammatical description can be
re-used without change.
This report may also be read as simply a linguistic description of Urdu morphology. By ‘linguistic’
description, we mean a description that a knowledge provider with training in linguistics would
expect, in which the morphology is described using such traditional linguistic terms as ‘allomorph’
and ‘morphosyntactic features.’
While Urdu is one of the world’s largest languages in terms of the number of native speakers,
published grammatical descriptions tend to leave certain issues unclear, and at times even disagree
among themselves. We have consulted a number of different grammars in writing this description;
we have reconciled those differences, and we have filled in gaps in coverage wherever possible
by consulting with Urdu linguists and native speakers. To the extent which we have succeeded
in these tasks, this report may be read as an improvement on those existing grammars, at least in
the realm of morphology.
In some cases, the differences among previous grammatical treatments of Urdu have turned out
to be ambiguities in how those grammatical descriptions were phrased. Such ambiguity is an
inherent drawback to writing in English (or in any other natural language), and it has the potential
to affect our own writing as much as anyone else’s. Nevertheless, we feel that we have succeeded in
avoiding much of the ambiguity of previous grammars, for two reasons. First, our formal grammar
was written more or less in parallel with our descriptive grammar. One effect of this is that the
author of the formal grammar frequently consulted with the authors of the descriptive grammar
about ambiguities or other difficulties with the description; in the process, the description was re-
written to remove ambiguities. Second, the formal grammar is (as far as possible) unambiguous,
and can therefore be consulted in the case of remaining ambiguities in the descriptive grammar.13
In order to make the formal grammar more accessible to linguists (particularly to linguists who are
not familiar with the XML notation), we intend in the future to produce a form of this document in
which the XML notation for the formal grammar has been converted to a notation more similar to
12
An electronic dictionary is normally a required resource. Fortunately, dictionaries are almost always more easily
obtained than grammars, at least grammars of the sort required for morphological parsing. There are two situations
where it might nonetheless be useful to do morphological parsing without a dictionary. One is where the target
language is more or less undescribed, but where there is a dictionary of a closely related language that can be searched
for cognate words, such as ‘haus’ and ‘house’ in German and English. The technology for this is not yet in place, but
could be developed. The other situation where a dictionary may not be needed is with proper nouns, for languages in
which names take inflectional morphology. Removing pieces of names that might be inflectional affixes may make a
name more recognizable or allow tracking references to names which differ by virtue of the inflectional affixes.
13
Of course, there is no guarantee that the formal grammar is correct. We expect that errors and inconsistencies will
show up as we implement the formal grammar in a parsing tool, and use this tool on Urdu texts. Any corrections will
be factored back in to our written grammars, both descriptive and formal.
6
Introduction
a traditional linguistic description. For example, the XML representation of inflectional affixation
in agglutinating languages uses a structure which (in somewhat simplified form) looks like the
following:
<Gr:PartOfSpeech name="verb'>
<Gr:affixSlots>
<Gr:InflAffixSlot id="slotCausative'>
<Mo:InflectionalAffix idref="afCausative'/>
<Mo:InflectionalAffix idref="afDoubleCausative'/>
</Gr:InflAffixSlot>
<Gr:InflAffixSlot id="slotNonfinite'>
<Mo:InflectionalAffix idref="afImpPtcplMsSg'/>
<Mo:InflectionalAffix idref="afImpPtcplMsPl'/>
<Mo:InflectionalAffix idref="afImpPtcplFmSg'/>
<Mo:InflectionalAffix idref="afImpPtcplFmPl'/>
<!-- more affixes here-->
</Gr:InflAffixSlot>
<!-- more slots here-->
</Gr:affixSlots>
<Gr:affixTemplates>
<Gr:InflAffixTemplate>
<Mo:refSuffixSlots>
<Mo:SuffixSlot name="Qualifier' idref="slotCausative'/>
<Mo:SuffixSlot name="Case' idref="slotNonfinite'/>
</Mo:refSuffixSlots>
</Gr:InflAffixTemplate>
</Gr:affixTemplates>
</Gr:PartOfSpeech>
For a linguist, a more useful (and more readable) display of this XML structure might be the
following table, in which the slots have been treated as columns in a table of affixes, and the
individual affixes are displayed as a pairing of a gloss and a form,14 rather than a reference to the
definition of the affix elsewhere (the ‘idref’ in the above XML code):
-tī(ṁ )
“ImperfectFemPlPtcpl”
(etc.)
14
For the sake of simplicity, the form is displayed here transcribed in a Latin writing system, rather than in the native
Urdu script.
7
Introduction
We emphasize that this is a matter of how the XML structure is displayed, not a change in the
underlying XML. This particular step (the conversion process from our XML-based grammar to a
display in the form of tables or other forms familiar to linguists) is not implemented yet. However,
the use of an XML formalism for both the descriptive and formal grammars means that when
the display mechanism is programmed, we will be able to produce versions of this grammar as
PDFs and other formats which incorporate the new display, without changing the underlying XML
documents.
In keeping with the purpose of this project, this grammar covers the morphology of standard
spoken and written Urdu, especially the inflectional morphology. Where possible, we have docu-
mented variation in Urdu, both in its spoken form and in its spelling. However, this report does
not constitute field work on dialectal variation, nor of its use by non-native speakers. In particu-
lar, while Urdu is the official language of Pakistan, it is not the first language of many inhabitants
of that country. There is likely to be considerable variation among non-native speakers, including
the use of non-Urdu words and code switching.
This grammar does not cover the phonology of Urdu, except insofar as some of the allomorphs
can be seen to be derived by phonological rules.15 That is, the grammar covers what is sometimes
called morphophonemics, but not phonemics. The grammar also does not cover syntax, except at
the boundaries of morphology, such as the uses of various case markers.
We suspect that there is considerable variation in spelling, in part because there are a number
of letters in the Perso-Arabic script used to write Urdu, which are homophonous in present day
spoken Urdu.16 It seems likely that writers of Urdu (including those for whom it is their first
language) err in their choice among homophonous letters, but this is not well documented in the
literature, and we have not studied it in the corpora to which we have access. This could obviously
be important in the application of computer processing to printed texts.
In this section we describe some enhancements which could be done to make this grammar usable
for other purposes.
The original objective of the project was to create portable grammars that could be converted into
morphological parsing engines and used to assist linguists in doing dictionary lookup, as described
in the previous section. However, while we were writing the Bengali and Urdu grammars, it
became apparent that the grammars could serve additional purposes as well. One such purpose
would be to serve as an explanation of the meaning of grammatical affixes: in essence, a ‘help’
15
For example, in- and im- are two allomorphs of an English prefix. Im- occurs before words beginning with ‘b,’ ‘p,’
and ‘m’ like impossible while in- occurs elsewhere like inappropriate.
16
Examples of homophonous letters or letter sequences in English are ‘k,’ ‘ck,’ and the “hard” ‘c.’
8
Introduction
document. In many languages (including Urdu), affixes provide essential clues—sometimes the
only clues—to who is the actor and who or what is acted on; or whether an action has happened
in the past, or is expected to happen in the future, or is even something that could have happened
but didn’t (a counterfactual). While the parser supplies a label for these affixes, the meaning of
the label may not be immediately clear to the user without some further description.
In addition, affixes can be used in what would otherwise be inappropriate situations to produce a
certain effect (such as the use of a feminine gender as a kind of diminutive). Knowledge of facts
like these is obviously important when trying to understand the meaning behind the meaning of
a text.
As with many of the applications described here, a user who is highly fluent in the target language
would probably not need help interpreting this information. But a user who did not know the
language well might be puzzled by the descriptions which the morphological parser attaches to
affixes. It would be inappropriate to attach paragraph-sized or page-length descriptions to every
occurrence of an affix, but it would not be difficult to attach a link from each occurrence to such
a description, so that a user could click on the link and be given the grammatical description.
Affixes, like any other fact about a language text, can also provide important information about
the speaker. For example, where there is dialectal variation in language usage, it may be used
to determine the origin of a speaker. As mentioned above, we have not extensively documented
such dialectal variation, although this could be done in future work.
In summary, in addition to the aid to the linguist provided by automated word analysis, we might
use our grammar (suitably rephrased) as the basis for a help document, with the information about
how affixes are used linked to the affixes found by the parser. Such an application would require
further programming effort.
A morphological parser constitutes a spell checker, but it does not by itself constitute a spell
corrector. However, it is possible to build a spell corrector on top of a morphological parser. As we
have worked on the Bengali and Urdu grammars, it has become clear to us that for these languages
at least, spelling is probably as much of a hindrance to dictionary lookup as the morphology.
(Other languages will vary; at one extreme are languages like Vietnamese, where there is no
morphology to hinder dictionary lookup at all; at the other extreme are languages of the Semitic
family, such as Arabic or Amharic, and languages of the Philippines, such as Tagalog or Ilocano,
where mastery of the morphology—or the use of a morphological parser—is absolutely required
in order to find the dictionary lookup form of an inflected word.)
Spelling errors fall into one of several different classes, and the methods that could be used to do
spell correction differ for each of these. The first class of errors is caused by alternative ways of
writing a certain sound. English is notorious for this. One example is the sound usually written
with the letter ‘f,’ as in ‘fish.’ The sound is sometimes written with a ‘ph,’ as in ‘philosophy,’ and
rarely with a ‘gh,’ as in ‘enough.’ Fortunately for users of English dictionaries, the cases where
such alternative spellings happen tends to be after the first few letters of the word (‘philosophy’
is an exception to this tendency). The result is that if you can figure out the first few letters of
the word, you are generally at about the right place in the dictionary, and a little scanning of the
page will find the right spelling. This tendency does not, however, help the user of the Bengali or
Urdu dictionary, where the spelling difference is equally likely to be in the first letter of the word.
9
Introduction
When combined with the fact that the writing systems allow for up to four variant spellings of
certain sounds, this makes finding a word in a printed Urdu or Bengali dictionary an exercise in
frustration.
For lookup in electronic dictionaries, the use of “wild cards”—a symbol such as an asterisk standing
for any letter—can be of some help. However, since such a wild card can match any letter, whereas
the actual ambiguity in the spelling of a particular sound is only one of a few letters, the use of
wild cards may return far more results than the desired ones. Imagine for example that one was
unsure whether the English word ‘fan’ was spelled with an ‘f’ or a ‘ph’; using a wild card for the
first letter would return not only ‘fan,’ but also ‘ban,’ ‘can,’ ‘man,’ ‘ran,’ and so forth. Combined
with other difficulties in spelling of Urdu (and to a lesser extent, Bengali) words, the result can be
that lookup in an electronic dictionary of words for which one is uncertain of the spelling can be
frustratingly difficult.
Knowing something of the spelling of Urdu or Bengali, it is possible to describe the potential errors
of this sort using “regular expressions17 .” Using the English ph-f case for purposes of illustration,
instead of looking up ‘*an,’ the regular expression ‘(f|ph)an’ can be used to reduce the ambiguity
to the exact two possibilities.
One could train people doing dictionary lookup to build such regular expressions for their search
queries. However, we suspect that not all users of electronic dictionaries will be comfortable with
doing this, and one can imagine that users would tend to revert to the simpler, but clumsier, use
of wild cards. It can also be difficult to think of all possible ambiguities. Fortunately, since the
ambiguous letters are known ahead of time, it would be possible to pre-program these, so that
if the user typed in a word with ambiguities, the program could create the regular expression
accounting for all possible ambiguities, and this regular expression would be used for dictionary
lookup (perhaps after being morphologically parsed).
Of course this should not be done for all users; someone who is knowledgeable in Urdu spelling
will probably type the word right the first time, and would not appreciate homophones that might
show up , just as a good English speller would not want to be told that the word “their” that they
typed in might be “they’re” or “there.”
A second class of spelling error arises from the fact that some sounds are more difficult to dis-
tinguish than other. This is the sort of error that non-native speakers are more likely to make,
although in the presence of noise even native speakers can mistake the English word “pill” for
“bill,” say. The possible solutions to this problem are exactly the same as for the first kind of
spelling difficulty, namely the use of wild cards, human-generate regular expressions, or regular
expressions created by a program which contains knowledge of the easily confusable pairs (or
perhaps multiples) of sounds. Some research could be done to see what these pairs are, although
for Urdu it seems likely that the dental vs. retroflex distinction can be difficult for native speakers
of English to learn, and perhaps also the long vs. short vowel distinction. Again, the user should
have the choice of seeing this sort of help or not.
The third class of spelling errors is random error, such as hitting the wrong key, or typing two keys
in the wrong order. Unlike the first two classes of errors, it is in general not possible to predict
these errors (although typing a key next to the desired key might be more frequent than typing
a key that is distant from the desired key). Wild cards and regular expressions are therefore of
little help here. Instead, the lookup program must allow for random errors in a kind of “fuzzy”
17
Regular expressions are a mathematically well-defined way of expressing certain kinds of ambiguity; they have a
computationally tractable implementation.
10
Introduction
lookup, and it must penalize such errors, since otherwise too many candidate corrections will be
generated.
The Stuttgart Finite State Tools (SFST) which are being used as the parsing engine in this project
does have exactly such a fuzzy lookup capability, so it would not be difficult to incorporate this
into the parser. This does however introduce the potential for increased memory load, an issue
which would need to be studied before this idea was implemented.
In sum, we believe that spelling correction is an important technology that should be developed
in conjunction with the morphological lookup capabilities already being built in this project.
As we look to the future, it is clear that Urdu is only one of hundreds of languages for which one
might want to build morphological parsers.18 All of these, we believe, can benefit from grammar
writing using the dual-use framework we have developed for Urdu. However, it is obvious that
we cannot write that many grammars ourselves. There are two ways that our work could be
leveraged, so as to make grammars of a large number of languages, and tools built on those
grammars, available: by having it serve as a model or template for other grammar writers; and
by automatically or semi-automatically adapting the formal grammar of one language to another
language. The two sub-sections below discuss each of these approaches in turn.
The traditional way to produce morphological parsers is to rely on highly trained linguists and
computational linguists to learn the programming language for some morphological parsing en-
gine (or to write one’s own parsing engine), learn the grammar and perhaps the writing system
of the target language, and then use the former knowledge to encode the latter knowledge. An
obvious impediment to this approach is that it is difficult to find one person who combines all
these skills. Another difficulty, discussed above, is that parsing engines tend to be replaced with
newer and better engines after a few years, rendering the parser that was built with so much
expert effort obsolete.
The dual use grammar method which we have developed in this project provides a way to avoid
the first problem: to the extent that the prose and descriptive grammars are separable, they can be
written by two different people who bring two different skill sets: one, knowledge of the grammar
(and writing system) of the target language; the other, experience in computer programming. It
is, we believe, easier to find two different people (or perhaps two teams of people), one with
each of these skill sets, than it is to find one person with both skills. We have in fact employed
this division of labor in both the Bengali and the Urdu grammars, and it has become clear that
the dual use grammar approach makes it easier to build teams who can construct grammars and
morphological parsers.19
18
There are in the neighborhood of 7000 languages in the world today (http://ethnologue.org is the standard refer-
ence on languages of the world). Of these, perhaps 1500 to 2000 are written languages, and probably the majority of
these have non-trivial inflectional morphologies. Over 300 languages have at least a million speakers.
19
The two grammars must still be written collaboratively, which calls for a close working relationship between the
descriptive grammar writer and the formal grammar writer. While the authors have worked in nearby offices, we
believe that this working relationship can probably be more remote; e.g., it might be mediated by email. We have not
done this experiment, although we believe it is worth trying.
11
Introduction
In addition, we have provided a more durable way to write formal grammars, one which will not
be made obsolete by the introduction of a new and better parsing engine.
It may be possible to further reduce the expertise needed to write grammars, if new grammars
can be modeled after existing grammars. To some extent, this Urdu grammar re-uses the model of
our earlier Bengali grammar, although we have introduced some new techniques the second time
around. In particular, we have written the Urdu grammar with a non-technical (or ‘non-scientfic’)
linguist in mind, in addition to a technical linguist, with the idea that the resulting grammar
may serve as a grammar help as described above, assuming some we are tasked to develop the
infrastructure for this.
We suspect that using our grammars as models would work best if the new grammars were being
written for languages related to the ones we have described. But the use of model grammars may
prove useful for unrelated languages as well.
As a further extension of our division of labor, we foresee that the production of the descrip-
tive grammar component—and possibly of the formal grammar—could be outsourced. While
outsourcing is not within the current scope of TTO 1308, but we would welcome ideas on how
it might be done. Initial experiments should probably involve collaboration with either nearby
linguists and computer programmers, or with individuals who have experience in computational
linguistics already. Later experiments might involve teaming up linguistics departments in the
countries where the languages are spoken (who would write the descriptive grammars), together
with computer science departments, perhaps at universities in the same countries (who would
write the formal grammars).
Some testing and debugging tools should also be developed for verifying the grammars using
parsers derived from the formal grammars. We have written some such tools in the process of
building the Bengali parser, but more could (and, we believe, should) be done so that the example
sentences and paradigm tables of the descriptive grammar can be better used as a source of test
cases. The present system lends itself to extracting such test cases, but does not make it easy to
compare the analyses returned by the parser with the analysis in the text. We have ideas on how
this two-way communication could be improved.
Rather than having linguists and computer programmers write grammars, another approach would
be to create a computer program that could automatically adapt an existing formal grammar
to work for a third language, related to the initial target language. (A computer could not be
expected to adapt a descriptive grammar, since that would require understanding of an English
grammatical description, a task which is well beyond current capabilities.) This task might be
done with various sorts of resources in the third language: corpora, bilingual corpora in the third
languages and English, bilingual corpora in the third language and the initial target language
(such as Bibles), dictionaries of the third language, etc.
Automatic grammar adaptation was one of the goals of the predecessor project to this one (TTO308).
However, due to lack of sufficient personnel this goal could not be pursued.
12
Introduction
20
The DocBook Literate Programming extensions can be downloaded from
http://docbook.sourceforge.net/release/litprog/current/dtd/ldocbook.dtd.
21
For many of the same reasons, international standards committees such as ISO are developing standard ways
of representing the contents of computer-readable dictionaries, representations which abstract away from particular
database standards, and which generally use XML for interchange purposes.
22
This is not to say that support for the display of Unicode is completely satisfactory. Urdu is standardly written in
the Nasta’liq form of the Arabic script. With few exceptions, fonts for the Arabic script use the Naskh form rather than
the Nasta’liq form, and most computers will display Urdu text with a Naskh font. While the result is technically the
same, it can be quite difficult for people who are used to one form of the Arabic script to read the other. There may also
be significant rendering differences among different platforms (such as Windows vs. Macintosh). Because Nasta’liq
is the standard for Urdu, we have used a Nasta’liq font to render this document as a PDF. Readers should remember
13
Introduction
that it will look quite different if it is rendered on-screen (e.g. as HTML), unless a Nasta’liq font is used. (Even with a
Nasta’liq font, rendering of Urdu text can be quite challenging.)
Display issues on non-compliant computers can be resolved by use of the PDF format. Fortunately, the DocBook XML
format we use for editing is easily converted to PDF.
14
The Urdu Language
Chapter 2
2.1 Background
Urdu is an Indo-Aryan language of South Asia, the official language of Pakistan, and one of 23
official languages in India. In addition to being spoken in India and Pakistan, Urdu is also spoken
in the South Asian countries of Bangladesh, Afghanistan, and Nepal, as well as by Muslim members
of the South Asian diaspora now living in the Middle East, Europe, the United States, and Canada.
It is the first language of at least 60.5 million speakers, with another 40 million or more who speak
it as a second language.1
The history of Urdu, its dialectal complexity, and its linguistic and cultural status, particularly in
relation to Hindi, are all fascinating and complicated issues that this grammar can barely touch
upon; a brief account will have to do.
2.1.1 History
Urdu and Hindi both descend from what was the lingua franca for the Mughal court2 of Delhi,
India, and its environs and adherents. That language was a standardized amalgam of the regional
dialects of the time, and as such was also influenced by the Kauravi, Hariyanvi, Panjabi, Rajasthani
(Mewati), and Braj languages, among others, in addition to Arabic, Persian, and Turkish elements
brought to the region by the Mughals. Under the British Raj,3 the language, known variously as
Urdu,4 Hindi, Hindustāni, and Kʰaṛī Boli, gained further ground as the standard common language
of the region.
Nowadays this common language is most often called Kʰaṛī Boli and is generally considered to
consist of four main varieties: Urdu, Hindi, Dakʰini, and Rekʰta. Dakʰini is the southern variety,
spoken in and around Hyderabad in the Indian state of Andhra Pradesh, as well as by Muslims in
the other towns and cities of India’s Deccan region and in Mumbai. Rekʰta is the literary, poetic,
1
Summer Institute of Linguistics Ethnologue.
2
Early sixteenth to mid-nineteenth centuries.
3
Mid-nineteenth to mid-twentieth centuries.
4
Short for ز ناردو/zabān-e-urdū-e-muallā/ language of the Exalted Camp. اردو/urdū/ means camp and in this usage
referred to the Imperial Bazaar.
15
The Urdu Language
and formal variety of Urdu, strongly influenced by Persian. Within each of the first three types,
there exists widespread dialectal variation.
Urdu shows extreme register5 variation that ranges from the highly Persianized language of Urdu
ghazals (a poetic form) and other literature to the highly Anglicized Urdu of the upper class
and media. Additionally, although it is the official language of Pakistan, and one of the official
languages of India, many speakers of Urdu do not speak it natively. In Pakistan especially, it is
learned as a second language after a local language (Punjabi, Sindhi, Balochi, etc.). This can cause
wide dialectal variation in phonology and grammar and therefore lead to problems for the linguist,
who must take care to be aware of the socio-cultural background of any informant. It is common
for these related languages to share lexemes, but to assign them different genders; for example,
native Punjabi or Sindhi speakers often change the genders of Urdu nouns: the word د/dahi/
yogurt is masculine in Urdu, but many Punjabis and Sindhis treat it as feminine. An additional
complication arises from the emigration, half a century ago, of the muhājirs, the Muslims who left
what is now India to settle in Pakistan after Partition.6
Furthermore, English is now exerting a powerful influence on many Urdu speakers, particularly
those of the upper classes. As is the case all over South Asia, in many people’s eyes, education
and class are measured by how much English one knows. This is partly a carryover from the days
of Colonial India, and also a result of increased globalization and work or education abroad by
Pakistanis. Bollywood in particular is becoming increasingly Anglicized, to the extent that films
now carry titles such as “Jab We Met” When We Met and “U, Me, aur Hum” You, Me, and Us.
These films, though produced in India, carry huge cultural weight in Pakistan as well and thus
influence the younger generation. Additionally, there is even greater exposure to Western media
due to satellite TV and to the preference for English vocabulary and Latin transcription of both
Urdu and English words, so that advertising and programming can have appeal on both sides of
the border.
The biggest register gap may be tied to age. Some members of the older generations tend to use
highly Persianized Urdu expressions, with much more Perso-Arabic vocabulary, while younger
generations will freely use English lexemes and even English phrases mid-speech. This style of
speech more closely reflects code-switching phenomena rather than borrowing (which involves
the occasional use of single words or phrases from the other language), usually with Urdu acting as
the matrix language and English filling in lexical gaps. Code-switching, the continual intermixing
by bilingual speakers of more than one language within single utterances or sentences, is a sign
of even heavier linguistic influence than mere borrowing.
As much as possible, this grammar has tried to document the realities of spoken Urdu while also
acknowledging its written forms. Most of our example sentences, therefore, attempt to reflect
natural speech rather than writing.
5
A register of a language is a version of it that a speaker uses for a particular situation; e.g., writing versus speaking,
or formal versus informal. A dialect is a variety of a language that one geographic, ethnic, socio-economic, etc., group
speaks as opposed to another group, like London English versus New York English, or “upper class” (New York) English
versus “blue collar” (New York) English.
6
In 1947, what had been British-ruled India was divided into the modern states of India and Pakistan.
16
The Urdu Language
The relationship between Hindi and Urdu is best summarized by Colin Masica:7
At the colloquial level, and in terms of grammar and core vocabulary, they are virtu-
ally identical; there are minor differences in usage and terminology (and customary
pronunciation of certain foreign sounds), but these do not necessarily obtrude to the
point where anyone can immediately tell whether it is ‘Hindi’ or ‘Urdu’ that is being
spoken. At formal and literary levels, however, vocabulary differences begin to loom
much larger (Hindi drawing its higher lexicon from Sanskrit, Urdu from Arabic and
Persian), to the point where the two styles/languages become mutually unintelligible.
To the ordinary non-linguist who thinks, not unreasonably, that languages consist of
words, their status as different languages is then commonsensically obvious, as it is
from the fact that they are written in quite different scripts (Hindi in Devanagari and
Urdu in a modified Perso-Arabic). . . Many complex social and political forces. . .
have conspired to pull the two ‘styles’ ever further apart. Their identity as separate
languages may now be regarded as a cultural fact, however anomalous linguistically.
Urdu is written primarily in a script borrowed from Arabic via Persia. Like those scripts, and
unlike the Devanagari script in which Hindi is written, the Urdu script is an abjad: only conso-
nants are represented in normal orthography. Long vowels are indicated through special uses of
consonantal characters, and may be represented in different ways in different parts of the word;
short vowels may be indicated by diacritics, but these are generally omitted (see discussion below
for exceptions). Urdu also employs digraphs-- combinations of characters which represent single
sounds-- in writing vowels and aspirated consonants.
Like Arabic, Urdu is written from right to left. Numerals, however, are written left to right, as in
the Brāhmī-derived Indic scripts. The script is cursive, and letters may have up to four distinct
forms, or allographs: the independent form, which is unconnected to other letters; the initial form,
connected only on the left; the medial form, connected on both sides; and the final form, connected
only on the right. (Unicode rendering automatically displays the proper positional form.)
Two types of Arabic calligraphy have been widely adapted for typography: Naskh and its deriva-
tive Nasta’līq. This grammar follows the more common Urdu usage in employing Nasta’līq script.
Urdu text may be rendered into Roman letters through transcription or transliteration. Transliter-
ation is one-for-one mapping of the Urdu characters (or character combinations) into correspond-
ing English characters (or combinations); its goal is to accurately represent the spelling of the
language. Transcription, on the other hand, maps the sounds of the Urdu language; its goal is to
accurately represent how the language is pronounced, not how it is written.
7
The Indo-Aryan Languages, pp. 27-30.
17
The Urdu Language
Because of the redundancy of the Urdu script (which possesses four distinct characters for the
sound /z/, three each for /s/ and /h/, etc.), no simple transliteration can preserve Urdu orthogra-
phy solely through the letters of the English alphabet; diacritics or other non-alphabetic characters
are required. Accordingly, we have chosen to use a broad phonemic transcription, rather than a
true transliteration. Urdu examples are transcribed in a version of traditional Urdu romanization,
slightly modified to more closely adhere to the International Phonetic Alphabet (IPA).
Where an example requires Urdu orthography, we give it in Urdu script, interlinear with its Ro-
manization.
Vowel length is distinctive in Urdu, though for some vowel pairs (long and short /i/, /e/, /o/,
and /u/), what was formerly only a length distinction has become a distinction of quality as well.
Long and short vowels are represented differently in Urdu orthography.
The short vowel signs (zabar, zēr, and pēš ), when written over or under a consonant, indicate
that the consonant is followed by the short vowels [a], [i], and [u] respectively. These signs are
seldom seen outside of dictionaries and instructional materials, and we do not use them in our
Urdu examples, but they are listed here for completeness:
Table 2.1 shows short vowel signs. For syllable-initial short vowels, the vowel sign is appended
َ◌8 zabar a a a
◌ِ zēr i i i
ُ◌ pēš ʊ u u
to a preceding alif. (The alif is commonly used without the diacritic, with context making it clear
which vowel is intended.) For example:
ُ
اس/us/ that
ِاس/is/ this
Initial vowels may also be indicated by ‘ عain (also spelled c ain); this character represents a pha-
ryngeal fricative in Arabic, a sound Urdu lacks. For example:
8
The small circles in this and subsequent tables stand in for the characters to which the diacritics are attached.
18
The Urdu Language
/alam/ flag (note that this spelling is ambiguous; it may also be read ِ /ilm/ knowledge.)
Initial [ ī ] (long [i]), [ē] (long [e]), and [æ] (as in English ash) are written with ایalif + yē. For
example:
ا ن/īmān/ faith
ا/ēk/ one
ا/æsā/ like this
Word-medially, they are represented by یyē alone. In final position, [ ī ] is represented by the
first final shape of yē ( یcʰōt ̣ī yē) and [e] and [æ] by the second ( ےbaṛī yē). When either form
of yē is used to represent [ ī ] or [æ] following another vowel, a hamzāh precedes the yē. Hamzāh̄
may also be placed directly over a vāū representing [ū] or [ō], in the same circumstances. Vowel
sequences beginning with [ ī ] generally do not take hamzāh.
Table 2.2 shows the hamzāh.
Initial [ō], long [ū], and [ɔ] are written with alif + وvāū.
اوڑ/oṛʰnā/ to drape (over the body)
اوﻻد/ɔlād/ children
اور/ɔr/ and
اوب/ūpar/ above
19
The Urdu Language
ڑا/pakɔṛā/ fritter
@ را/pūrā/ whole, entire
د/dʰōbi ̄/ washerman
Rarely, diacritics are used to distinguish among the three: pēš over the alif or the preceding
consonant for [u:], zabar for [ɔ]. There is no sign to specify [ō]. (Compare to front vowels in yē.)
َ
اور/ɔr/ and
ُ
اوب/ūpar/ above
Vāū, standing alone, represents short /u/ ([ʊ]) in two common words, دxud self and ش/xuš/
happy. In words of Persian origin, initial [xw] has been simplified to [x]; thus, in words beginning
xe-vāū-alif, the vāū is silent. This grammar does not transcribe this silent vāū.
Vāū after xe with no following alif is given its usual vocalic values.
Before an /h/ (written with either cʰot ̣ī hē or baṛi ̄ hē), written /a/ and /i/ are realized as [æ]; /u
+ h/ is realized as [a]. The /h/ in these cases is silent.
ا/æhmad/ Ahmed
اقa ا/æhtirāq/ combustion
ر9 /bahūr/ seas
We transcribe these words as they are spelled in Urdu, with the silent /h/ in place.
20
The Urdu Language
Similarly, before ‘ain-- /a/ lengthens to [ā], /i/ is realized [æ], and /u/ surfaces as [o], while the
‘ain itself is unpronounced.
> /tālīm/ education
اع ال/ætidāl/ abstinence
@ /tavaqqo/ hope
We transcribe words with these mutations as they are pronounced; silent ‘ain is omitted.
Word-finally, vowel nasalization is indicated by nūn ɣunnah: the letter nūn without its dot.
Table 2.3 shows final nasal vowel indicator nūn ɣunnah.
ب/mæṁ / I
Medially, nasal vowels are written with regular dotted nūn.
6 دا/dāṁ t/ tooth
We transcribe all nasal vowels, medially and finally, in the same way, with /ṁ /.
2.2.3.1 Transcription
The transcription system used in this grammar departs both from the IPA and from the tradi-
tional romanization of Urdu. Our intention is to create a system that accurately reflects Urdu
pronunciation while preserving phonemic and morphological information.
In keeping with Indological tradition, we use /c/ and /j/ for the palatal consonants and /y/ for
the palatal glide; and we represent the retroflex sounds with underdot.
We use /v/ for consonantal uses of vāū, and gamma, /ɣ/, for the velar fricative.
وا ب/vālidæn/ parents
غ/bāɣ/ garden
21
The Urdu Language
This transcription does not directly represent the two rare diacritics tašdīd and jazm (also called
sukūn, after the equivalent Arabic character).
Table 2.4 shows the consonantal diacritics tašdīd and jazm.
geminate
ّ◌ tašdīd consonant - CC
(rare)
absence of
ْ◌ jazm - -
vowel (rare)
Tašdīd indicates that a consonant is geminated (doubled); we transcribe all geminate consonants by
repeating the letter or letter combination representing it, rather than by using a special character.
ّہ9 /baccah/ child
Jazm indicates the absence of a vowel (compare to the Sanskrit virama). It is rare in Urdu orthog-
raphy, and omitted entirely in our transcriptions, which represent all vowels.
2.2.4.1 Vowels
Because word-internal short vowels are not represented in the Urdu script,9 transcription or
transliteration from Urdu to Roman script can be problematic for non-fluent readers. The long
vowels vāū and yē also present difficulties because each one represents several different sounds in
Urdu.
Another common problem lies with dialectal differences in pronunciation of vowels; for example,
speakers of north Indian Urdu will often have a higher vowel ([ɛ], as in English egg) for the sound
that is traditionally transcribed as /ai/ than will Pakistani speakers, whose pronunciation is more
likely to be [æ], as in English cat. (Hence our use of /æ/, as noted above.) However, a detailed
account of dialectal differences in pronunciation is beyond the scope of this grammar.
2.2.4.2 Consonants
Consonants also present a challenge for non-fluent users of Urdu because there are several cases
of multiple representations for a single sound:
9
With the exception of reading primers for children or other pedagogical or explanatory materials that have been
deliberately designed for explicitness.
22
The Urdu Language
• /h/ can be spelled with حbaṛi ̄ he anywhere or ہcʰot ̣ī hē word-initially and medially.
As a matter of fact, even native Urdu speakers sometimes struggle with the multiple represen-
tations. A major reason for these orthographic complexities lies in widespread borrowing: the
letters ثse and ذzāl are only seen in words of Arabic or Persian origin, and the letters صsvād,
ضzvād, ظzoe, and طtoe are confined to Arabic loanwords.
In addition, speakers of English or other western European languages will often encounter diffi-
culty in distinguishing between retroflex and dental consonants, as well as between the four-way
distinction of phonation among obstruent (= stop) consonants, traditionally labelled as voiceless
unaspirated, voiceless aspirated, voiced unaspirated, and voiced aspirated.
Table 2.5 shows consonants; Table 2.6 shows initial and long vowels; Table 2.7 shows short vowel
signs and vowel combinations; Table 2.8 shows diacritics.
ب bē b b b
پ pē p p p
ت tē t̪ t t
ٹ t ̣ē ʈ ṭ ṭ
ث sē s s s
23
The Urdu Language
ج jīm ʤ j j
چ cē ʧ c c
خ xē x x x
د dāl d̪ d d
ڈ ḍāl ɖ ḍ ḍ
ذ zāl z z z
ر rē r r r
ڑ ṛe ̄ ɾ ṛ ṛ
ز zē z z z
ژ žē ʒ ž ž
س sīn s s s
ش šīn ʃ š š
ص svād s s s
ض zvād z z z
24
The Urdu Language
ط toē t̪ t t
ظ zoē z z z
vowel quality
ع ‘ain ‘, ‘a ‘, -
marker
غ ɣain ɣ ɣ ɣ
ف fē f f f
ق qāf q q q
ک kāf k k k
گ gāf g g g
ل lām l l l
م mīm m m m
ن nūn n n n
The formal grammar’s listing of these graphemes (and one morpheme boundary marker) appears
below.
25
The Urdu Language
ِا i (ɪ) i i
ِای i: ī ī
ُ
ا u (ʊ) u u
ُ
او u: ū ū
اے e e ē
او o o ō
أے ɛ ai æ
أو ɔ au ɔ
َ◌ zabar a a a
◌ِ zēr i i i
ُ◌ pēš ʊ u u
i (following
◌ cʰōt ̣ī yē with hamzāh i i
another vowel)
ɛ (following
ۓ baṛī yē with hamzāh ai æ
another vowel)
u:, o (following
ؤ vāū with hamzāh ū, ō ū, ō
another vowel)
26
The Urdu Language
final nasal
ں nūn ɣunnah ṅ or ṁ ṁ
vowel
ٔ◌ hamzāh hiatus - -
absence of
vowel
◌ْْ jazm - -
(normally
omitted)
geminate
◌ّ tašdīd - CC
consonant
<Ph:PhonemeSet>
<!-- Phonemes (or graphemes, really). Some of these are substrings of
others. Also, some are ambiguous as to consonant vs. vowel.
In particular, the Vau (')'و, Choti He (')'ہ, Ye (')'ی, and
Do Cashmi He (' )'ھcan function as both consonant and vowel.
27
The Urdu Language
28
The Urdu Language
<!--Diacritrics-->
<Ph:Phoneme id='phNunGunna'>
<!--Transliteration 'ṁ' (nasalization)-->
<Ph:Form spelling=''ں/>
<!--U+6BA-->
</Ph:Phoneme>
<Ph:Phoneme id='phHamza'>
<!--Transliteration '.' -->
<Ph:Form spelling=''ء/>
<!--U+621-->
</Ph:Phoneme>
<Ph:Phoneme id='phSukun'>
<!--Transliteration '-'. Sukun is the Arabic/ Unicode name;
in Urdu, this is called 'tashdeed'-->
<Ph:Form spelling='ْ'/>
</Ph:Phoneme>
<Ph:Phoneme id='phTashdid'>
<!--Transliteration 'CC' (geminate). Unicode calls
this by the Arabic term 'shadda') -->
<Ph:Form spelling='ّ'/>
<!--U+651-->
</Ph:Phoneme>
<!--Consonants-->
<Ph:Phoneme id='phAlif'>
<!--Transliteration '-' -->
<Ph:Form spelling=''ا/>
</Ph:Phoneme>
<Ph:Phoneme id='phBe'>
<!--Transliteration 'b' -->
<Ph:Form spelling=''ب/>
<!--U+628-->
</Ph:Phoneme>
<Ph:Phoneme id='phBhe'>
<!--Transliteration 'bʰ' -->
<Ph:Form spelling='' ھ/>
<!--U+628 U+6BE-->
</Ph:Phoneme>
<Ph:Phoneme id='phPe'>
<!--Transliteration 'p' -->
<Ph:Form spelling=''پ/>
<!--U+67E-->
</Ph:Phoneme>
<Ph:Phoneme id='phPhe'>
<!--Transliteration 'pʰ' -->
29
The Urdu Language
30
The Urdu Language
</Ph:Phoneme>
<Ph:Phoneme id='phBarihe'>
<!--Transliteration 'h' -->
<Ph:Form spelling=''ح/>
<!--U+62D-->
</Ph:Phoneme>
<Ph:Phoneme id='phXe'>
<!--Transliteration 'x' -->
<Ph:Form spelling=''خ/>
<!--U+62E-->
</Ph:Phoneme>
<Ph:Phoneme id='phDal'>
<!--Transliteration 'd' -->
<Ph:Form spelling=''د/>
<!--U+62F-->
</Ph:Phoneme>
<Ph:Phoneme id='phDHe'>
<!--Transliteration 'dʰ' -->
<Ph:Form spelling=''دھ/>
<!--U+62F U+6BE-->
</Ph:Phoneme>
<Ph:Phoneme id='phDDotal'>
<!--Transliteration 'ɖ' -->
<Ph:Form spelling=''ڈ/>
<!--U+688-->
</Ph:Phoneme>
<Ph:Phoneme id='phDDotHe'>
<!--Transliteration 'ɖʰ' -->
<Ph:Form spelling=''ڈھ/>
<!--U+688 U+6BE-->
</Ph:Phoneme>
<Ph:Phoneme id='phZal'>
<!--Transliteration 'z' -->
<Ph:Form spelling=''ذ/>
<!--U+630-->
</Ph:Phoneme>
<Ph:Phoneme id='phRe'>
<!--Transliteration 'r' -->
<Ph:Form spelling=''ر/>
<!--U+631-->
</Ph:Phoneme>
<Ph:Phoneme id='phRDotE'>
<!--Transliteration 'ṛ' -->
<Ph:Form spelling=''ڑ/>
<!--U+691-->
</Ph:Phoneme>
<Ph:Phoneme id='phRhe'>
31
The Urdu Language
32
The Urdu Language
<!--U+639-->
</Ph:Phoneme>
<Ph:Phoneme id='phGain'>
<!--Transliteration 'ɣ' -->
<Ph:Form spelling=''غ/>
<!--U+63A-->
</Ph:Phoneme>
<Ph:Phoneme id='phFe'>
<!--Transliteration 'f' -->
<Ph:Form spelling=''ف/>
<!--U+641-->
</Ph:Phoneme>
<Ph:Phoneme id='phQaf'>
<!--Transliteration 'q' -->
<Ph:Form spelling=''ق/>
<!--U+642-->
</Ph:Phoneme>
<Ph:Phoneme id='phKaf'>
<!--Transliteration 'k' -->
<Ph:Form spelling=''ک/>
<!--U+6A9-->
</Ph:Phoneme>
<Ph:Phoneme id='phKhe'>
<!--Transliteration 'kʰ' -->
<Ph:Form spelling='ھD'/>
<!--U+6A9 U+6BE-->
</Ph:Phoneme>
<Ph:Phoneme id='phGaf'>
<!--Transliteration 'g' -->
<Ph:Form spelling=''گ/>
<!--U+6AF-->
</Ph:Phoneme>
<Ph:Phoneme id='phGhe'>
<!--Transliteration 'gʰ' -->
<Ph:Form spelling=''گھ/>
<!--U+6AF U+6BE-->
</Ph:Phoneme>
<Ph:Phoneme id='phLam'>
<!--Transliteration 'l' -->
<Ph:Form spelling=''ل/>
<!--U+644-->
</Ph:Phoneme>
<Ph:Phoneme id='phMim'>
<!--Transliteration 'm' -->
<Ph:Form spelling=''م/>
<!--U+645-->
33
The Urdu Language
</Ph:Phoneme>
<Ph:Phoneme id='phNun'>
<!--Transliteration 'n' -->
<Ph:Form spelling=''ن/>
<!--U+646-->
</Ph:Phoneme>
<Ph:Phoneme id='phVau'>
<!--Transliteration 'β, u, o, ɔ' (sometimes as vowel) -->
<Ph:Form spelling=''و/>
<!--U+648-->
</Ph:Phoneme>
<Ph:Phoneme id='phChotiHe'>
<!--Unicode calls this the Heh Goal.
Transliteration 'h' (sometimes as vowel) -->
<Ph:Form spelling=''ہ/>
<!--U+6C1-->
</Ph:Phoneme>
<Ph:Phoneme id='phYeh'>
<!--Unicode calls this the Farsi Yeh; it is the same as the Choti Yeh
(not to be confused with the Choti He).
Transliteration 'y, ī, e, ɛ' (sometimes as vowel) -->
<Ph:Form spelling=''ی/>
<!--U+6CC-->
</Ph:Phoneme>
<Ph:Phoneme id='phDoCashmiHe'>
<!--Transliteration 'ʰ' (sometimes as vowel) -->
<Ph:Form spelling=''ھ/>
<!--U+6BE-->
</Ph:Phoneme>
34
The Urdu Language
</Ph:PhonemeSet>
<!--This grammar fragment contains some spelling rules. These are similar to
phonological rules, but apply to the spelling without making a difference
in the sound.
-->
<Ph:PhonologicalRule name="ruleGeminateYeh">
<!--When two Farsi Yehs come together, the first becomes a Yeh
with Hamza above. This applies in the nasal declension of adjectives,
and possibly elsewhere.
-->
<Ph:subrules>
<Ph:PhonologicalSubrule>
<Ph:input><Ph:refPhoneme idref="phYeh"/></Ph:input>
<Ph:output><Ph:refPhoneme idref="phYeHamza"/></Ph:output>
</Ph:PhonologicalSubrule>
</Ph:subrules>
<Ph:environments>
<Ph:Environment>
<Ph:rightContext>
<Ph:SequenceContext>
<Ph:SimpleContextTerminal>
<Ph:refBoundary idref="bdryMorpheme_bdry"/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref="phYeh"/>
</Ph:SimpleContextTerminal>
</Ph:SequenceContext>
</Ph:rightContext>
</Ph:Environment>
</Ph:environments>
</Ph:PhonologicalRule>
These are the abbreviations and symbols that appear in the interlinear text examples in this gram-
mar:
• *: ungrammatical form
• 1: first person
• 2: second person
35
The Urdu Language
• 3: third person
• ACC: accusative/dative
• ADJ: adjectival informant /vālā/
• CAUS: causative
• CP: conjuctive particle /kar/
• DIM: diminutive
• DIR: direct
• EMPH: emphatic
• ERG: ergative
• EXC: exclusive particle /hiˉ/
• F: feminine
• FUT: future
• INF: infinitive
• IMP: imperative
• INC: inclusive particle /bʰiˉ/
• IP: imperfective participle
• M: masculine
• NEG: negative
• OBL: oblique
• PER: perfective
• PL: plural
• POSS: possessive
• PP: perfective participle
• PRES: present
• PROG: progressive
• PST: past
• Q: question
• REL: relative
• SG: singular
• SBJV: subjunctive
• TAG: tag
36
Nouns
Chapter 3
Nouns
3.1 Inflection
Urdu nouns show gender (masculine and feminine) and number (singular and plural) and are
inflected for three cases: direct, oblique, and vocative. Categories such as possessive, direct and
indirect object, etc., are expressed by means of postpositions following the oblique case. (See
Chapter 5).
The formal grammar’s definition of the morphosyntactic features of nouns follows:
<!--The features used on nouns. This includes both the features that
nouns carry in the lexicon (gender and sometimes number), and the
features that they are inflected for (case and number).
-->
<Fs:ClosedFeatureDefn name='Gender' id='fdefnGender'>
<!--An inherent feature of nouns in the lexicon; adjectives agree
with the nouns they modify in this feature.-->
<Fs:SymbolicFeatureValueDefn id='fvMasculine' symbol='masculine'/>
<Fs:SymbolicFeatureValueDefn id='fvFeminine' symbol='feminine'/>
</Fs:ClosedFeatureDefn>
<Fs:ClosedFeatureDefn name='Number' id='fdefnNumber'>
<!--Sometimes an inherent feature of nouns in the lexicon, but
most nouns can inflect as either singular or plural.-->
<Fs:SymbolicFeatureValueDefn id='fvSingular' symbol='singular'/>
<Fs:SymbolicFeatureValueDefn id='fvPlural' symbol='plural'/>
</Fs:ClosedFeatureDefn>
<Fs:ClosedFeatureDefn name='Case' id='fdefnCase'>
<!--A feature which nouns inflect for.-->
<Fs:SymbolicFeatureValueDefn id='fvDirect' symbol='direct'/>
<Fs:SymbolicFeatureValueDefn id='fvOblique' symbol='oblique'/>
<Fs:SymbolicFeatureValueDefn id='fvVocative' symbol='vocative'/>
</Fs:ClosedFeatureDefn>
37
Nouns
3.1.1 Gender
Some nouns are recognizably either masculine or feminine by their endings, while the gender of
other nouns must simply be memorized. The first group is called “marked” and the second, “un-
marked.” There are therefore four categories of nouns: masculine marked, masculine unmarked,
feminine marked, and feminine unmarked. Inflection of a noun is based on which group it belongs
to; in other words, these four noun types are inflectional classes.
Note that although the discussion below describes noun classes in terms of their endings, those
endings cannot always be used to determine what class any particular noun belongs to, since
many unmarked nouns have endings identical to those on marked nouns, for example, the word
/hayā/ modesty is an unmarked feminine noun, even though alif ا/ā/ is the ending on most
marked masculine nouns. Similarly, د/dahī/ yogurt is masculine, even though it ends in /ī/,
which often marks feminine nouns.
Gender is arbitrary, except in living beings, where it can be determined by biological gender:
ں/māṁ / mother (feminine)
پ/bāp/ father (masculine)
a ڈا/ḍākt ̣ar/ doctor (masculine when referring to a man, feminine when referring to a woman)
There are also other unmarked masculine nouns that, although they may refer to female as well
as males, remain in the unmarked masculine inflectional class even when referring to females;
for example, دوس/dōst/ friend and ن/mihmān/ guest. But even though the noun is formally
masculine in those cases, any modifying adjectives or verbs that go with it will take feminine
endings:
There are also many animal nouns that have one grammatical gender, but can refer to a creature
of either sex; for example, ّ /makkʰī/ housefly (f.), meaning either a female or male fly; or ّ
/billī/ cat, which is also feminine but can refer to either a female or
ّ male cat when there is no
need to stress the cat’s sex, and has a masculine counterpart as well, /billā/, which is used when
the speaker wants to make a point that the cat is male.1
There are some instances in which the feminine gender is used to imply smallness of the object:
را /cʰurā/ dagger, large knife → ری /cʰurī/ small knife, table knife
ڈبّہ /ḍibbah/2 box → ّ ڈ/ḍibbī/ small box
There are also a few words whose gender varies with the dialect being used.
1
Compare the English parallel with cat and tomcat. The former can refer to either a female or male cat; the latter
only to a male. And although English nouns don’t have grammatical gender, the pronoun more often used for a cat, if
its sex is unknown and the speaker wishes to avoid using it, is she.
2
This word can also be pronounced /ḍabbah/.
38
Nouns
3.1.1.1 Masculine
All marked masculine nouns end in the sounds /ā/, /ayā/ , or, rarely, /āṁ /. But note that the con-
verse is not necessarily true: not all nouns that end in /ā/, /ayā/, or /āṁ / are marked masculines;
for example, ا/havā/ wind, air, /tatayyā/ red and yellow wasp, and ا ّ ں/ammāṁ / mother are all
in the unmarked feminine class. There are also final /ā/ nouns that fall in the unmarked masculine
class, even ones that refer to biologically male entities, such as را ہ/rājah/ or دادا/dādā/.3
Many Urdu nouns (and also adjectives; see Chapter 6) that end with the sound /ā/ are actually
spelled with a “silent” final cʰot ̣ī hē; for example, ّہ9 /baccah/ boy. The sound /ā/ at the end
of a word can be represented three different ways in the Urdu script so, orthographically, there
are three marked masculine noun endings representing the one long /ā/ sound: ا/ā/, ہ/ah/,
and بہ/ayah/. Many grammars omit the silent final cʰot ̣ī hē in their roman transcription, but as
mentioned in our transcription chapter, we will represent it in ours, for clarity, with a Roman /h/.
Many of these nouns with a final cʰot ̣ī hē are of Arabic origin; however, Urdu speakers are increas-
ingly spelling even Indic words with a final cʰot ̣ī hē instead of alif.
Examples of marked nouns:
5 ڑ/laṛkā/ boy
ّہ9 /baccah/ or 9/baccā/ (male) child
ہ6رو/rūpayah/ rupee, money
اں/kūāṁ / well
ر/gʰar/ house
م5/kām/ work
3.1.1.2 Feminine
Marked feminine nouns end in either ی/ī/ or /iyā/. But as with marked masculine nouns, not
all nouns with these endings fall in this class; for example, /pānī/ water and زی/jahāzī/ sailor
are both unmarked masculine nouns.
Examples:
ڑ/laṛkī/ girl
3
Although note that these words are inflected like marked masculine nouns in some dialects.
39
Nouns
ر/murɣī/ hen
ڑ/ciṛiyā/ bird
ب8/kitāb/ book
@ رت/ɔrat/ woman
a /mēz/ table
Although we said above that the gender of unmarked nouns must be learned, there are some
generalizations that can help in guessing to which gender a noun likely belongs. Remember,
however, that there are numerous exceptions: nouns with these endings do not necessarily follow
the gender patterns below. To be certain, one must check the dictionary.
Many of the suffixes listed below are also discussed in Section 3.3.
• و/-ō/ (often used to form women’s names in rural areas) (See Section 3.2.2)
40
Nouns
• و/-ū/
• و/-ō/8
• ب/-pan/ (noun-forming suffix) (See Section 3.2.4)
• /-pā/ (noun-forming suffix)
• ن /-(i)stān/ (Persian loanwords) (See Section 3.3.1.2)
Table 3.1 shows the paradigm of masculine marked nouns in ہ، ا/-ā, -ah/; Table 3.2 shows the
paradigm of masculine marked nouns in اں/-āṁ /; Table 3.3 shows the paradigm of masculine
marked nouns in بہ/-ayah/; Table 3.4 shows the paradigm of masculine unmarked nouns.
Singular Plural
Table 3.5 shows the paradigm of feminine marked nouns in ی/-ī/; Table 3.6 shows the paradigm
of feminine marked nouns in ا/-īyā/; Table 3.7 shows the paradigm of feminine unmarked nouns.
7
Note that unlike the other members of this list, گ ہ/gāh/ is not a suffix, but a compounding element.
8
This is most commonly a masculine suffix, but note that it can also occur at the end of women’s names. See list of
feminine suffixes above.
9
وں/-ōṁ / has been reported as a variant suffix in colloquial speech.
10
This pattern is rare and may only apply to one word, ہ6 روrūpayah ‘rupee.’
41
Nouns
Singular Plural
Singular Plural
Singular Plural
Direct ∅ ∅
Oblique ∅ وں/-ōṁ /
Vocative ∅ و/-ō/9
42
Nouns
Singular Plural
Singular Plural
Singular Plural
Direct ∅ ب/-ēṁ /
Oblique ∅ وں/-ōṁ /
Vocative ∅ و/-ō/9
43
Nouns
</Gr:PartOfSpeech>
The definition of suffix allomorphy based on the declension classes follows here; the declension
classes are assigned to nouns based on information in the dictionary (see Appendix B).
44
Nouns
<!--Masculine-->
<Mo:InflectionalAffix gloss='-MascSgDirect' id='afMascSgDirect'>
<Mo:AffixAllomorph>
<!-- -ā -->
<Ph:Form spelling='+'ا/>
<Mo:inflectionClasses>
<Mo:refInflClass idref='classNounMarkedV'/>
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Mo:AffixAllomorph>
<!-- -ah -->
<Ph:Form spelling='+'ہ/>
<Mo:inflectionClasses>
<Mo:refInflClass idref='classNounMarkedAh'/>
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Mo:AffixAllomorph>
<Ph:Form spelling='+'اں/>
<!-- -āṁ -->
<Mo:inflectionClasses>
<Mo:refInflClass idref='classNounMarkedAm'/>
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Mo:AffixAllomorph>
<Ph:Form spelling='+'بہ/>
<!-- -ayah -->
<Mo:inflectionClasses>
<Mo:refInflClass idref='classNounMarkedY'/>
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Mo:AffixAllomorph>
<Ph:Form spelling='+'/>
<!--Null allomorph-->
<Mo:inflectionClasses>
<Mo:refInflClass idref='classNounUnmarked'/>
45
Nouns
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvMasculine'/>
<Fs:SymbolicFeatureValue defn='fvSingular'/>
<Fs:SymbolicFeatureValue defn='fvDirect'/>
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
46
Nouns
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvMasculine'/>
<Fs:SymbolicFeatureValue defn='fvSingular'/>
<Fs:vAlt>
<Fs:SymbolicFeatureValue defn='fvOblique'/>
<Fs:SymbolicFeatureValue defn='fvVocative'/>
</Fs:vAlt>
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
47
Nouns
<Fs:SymbolicFeatureValue defn='fvPlural'/>
<Fs:SymbolicFeatureValue defn='fvDirect'/>
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
<!--Feminine-->
<Mo:InflectionalAffix gloss='-FemSg' id='afFemSg'>
<!--Syncretic in all three cases for all declensions -->
<Mo:AffixAllomorph>
<Ph:Form spelling='+'ی/>
<!-- -ī -->
<Mo:inflectionClasses>
<Mo:refInflClass idref='classNounMarkedV'/>
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Mo:AffixAllomorph>
<Ph:Form spelling='+ '/>
<!-- -iyā -->
<Mo:inflectionClasses>
<Mo:refInflClass idref='classNounMarkedY'/>
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Mo:AffixAllomorph>
48
Nouns
<Ph:Form spelling='+'/>
<!-- Null allomorph -->
<Mo:inflectionClasses>
<Mo:refInflClass idref='classNounUnmarked'/>
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvFeminine'/>
<Fs:SymbolicFeatureValue defn='fvSingular'/>
<!--Since this is used for all three cases, we don't mention
the case here. -->
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
49
Nouns
<Ph:Form spelling='+'وں/>
<!-- -ōm
̇ -->
<Mo:inflectionClasses>
<Mo:refInflClass idref='classNounUnmarked'/>
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvFeminine'/>
<Fs:SymbolicFeatureValue defn='fvPlural'/>
<Fs:SymbolicFeatureValue defn='fvOblique'/>
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
3.1.4.1 Masculine
3.1.4.1.1 Marked
Table 3.8 shows the paradigm of masculine marked nouns in ہ، ا/-ā, -ah/; Table 3.9 shows the
paradigm of masculine marked nouns in اں/-āṁ /. Table 3.10 shows the paradigm of masculine
marked nouns in بہ/-ayah/.
Singular Plural
50
Nouns
Singular Plural
Singular Plural
3.1.4.1.2 Unmarked
Singular Plural
11
The spelling of this form has the variants روand . رو
51
Nouns
3.1.4.2 Feminine
3.1.4.2.1 Marked
Singular Plural
3.1.4.2.2 Unmarked
Table 3.13 shows the paradigm of feminine unmarked nouns in ہ /xālah/ maternal aunt; Ta-
ble 3.14 shows the paradigm of feminine unmarked nouns in ب8 /kitāb/ book.
Singular Plural
3.1.5 Notes
• As in English, mass nouns (nouns for things that cannot be counted) such as ۓ/cāē/ tea,
/cīnī/ sugar, and so on, do not have plural forms, except when talking about, say, countable
units or measures of tea, sugar, etc.; for example, in the context of serving, as in I would like two
sugars in my tea, or Three coffees, please.
• Because most /ī/-final nouns are feminine and most /ā/-final nouns masculine, native speakers
tend to assign gender to foreign loanwords in accordance with that tendency.
52
Nouns
Singular Plural
3.1.6.1 Vocative
The vocative case is used when the speaker is addressing the person (or, rarely, thing) that the noun
names; hence, proper nouns and kinship terms are the nouns that most often take the vocative
case. A noun in the vocative can be preceded by ے/ē/, و/ō/, ا ں/amāṁ /, or ا/ajī/. See above,
Section 3.1.2, for the vocative endings, and Section 3.1.4 for examples of nouns in the vocative
case.
3.1.6.2 Direct
The direct case is called the “nominative” case in some grammars. It is the case used for subject
nouns and pronouns that are not followed by postpositions. The direct case can also be used with
direct object third person pronouns whose referents are inanimate, rather than the more usual
oblique + postposition:
See above, Section 3.1.2, for the direct endings, and Section 3.1.4 for examples of nouns in the
direct case.
53
Nouns
3.1.6.3 Oblique
The oblique case is used in two situations: when a noun is followed by a postposition, and when a
noun is being used adverbially, either in a locative or a temporal sense. Direct and indirect objects
are indicated by a postposition.
1. Noun + postposition:
(3.4) ے ر بa
mēr-ē gʰar meṁ
my-OBL house.OBL in
‘in my house’
(3.7) گۓ 9
See above, Section 3.1.2, for the oblique endings, and Section 3.1.4 for examples of nouns in the
oblique case.
Urdu has several suffixes that can be added to nouns, verbs, or adjectives to form new nouns. Most
are not productive, but some are, as indicated below. See Section 3.3 below for more examples
54
Nouns
3.2.1 وٹ/-vaṭ /
• Not productive.
• Examples:
د/dikʰānā/ to show > د وٹ/dikʰāvat ̣/ show, appearance
6 /banānā/ to make, build > وٹ6 /banāvat ̣/ form, shape, artificiality
7 /sajānā/ to decorate > وٹ7 /sajāvat ̣/ decoration, adornment
3.2.2 ؤ/-ō/
3.2.2.1 Masculine
• Not productive.
• Examples:
9 /bacānā/ to save, rescue, cause to escape > ؤ9 /bacāō/ defense
/bahānā/ to cause to flow > ؤ /bahāō/ flow
3.2.2.2 Feminine
• Examples:
رو /bahār-ō/ nickname for Bahar
ب/sab-ō/ nickname for Sabeen
55
Nouns
3.2.3 ی/-ī/
• Not productive.
• Examples:
ڑا/laṛānā/ to cause people to fight > ڑا/laṛāi/ war, battle
بڑ/paṛʰānā/ to teach > بڑ/paṛʰāi/ studying
• Example:
ا/banvānā/ to get something made > ا/banvāī/ cost of construction
• Examples:
ّ ا/accʰā/ good > ّ ا/accʰāī/ goodness
@ /bolnā/ to speak > @ /bolī/ speech; bid; dialect, language variety
56
Nouns
• Such words are actually primarily adjectival, with the meaning from (that place), although they
can also be used as nouns, glossed as above and in the examples below. (See Section 6.1.2.3.1
for their adjectival origin.)
• Examples:
وری /pešāvarī/ resident of Peshawar
و /hindūstānī/ Indian
3.2.4 ب/-pan/
• Not productive.
• Examples:
6 /nīlā/ blue > ب6 /nīlāpan/ bluishness
ّہ9 /baccah/ child > ب9 /bacpan/ childhood
3.2.5 /-haṭ /
• Not productive.
• Examples:
/ciknā/ oily, greasy > /ciknāhat ̣/ greasiness
3.2.6 ن/-n/
• Not productive.
• Examples:
/mālī/ gardener (male) > /mālan/ gardener (female)
ری /bʰikārī/ beggar man > رن /bʰikāran/ beggar woman
57
Nouns
When using words of Persian or Arabic origin, Urdu speakers have the option, with most of them,
of using either Urdu affixes or the morphology of the originating language. Using Persian or
Arabic morphology gives the language a higher register; i.e., makes it sound more formal and
literary. Borrowed nouns to which the original Arabic or Persian plural (or dual, in the case of
Arabic) morphology has been affixed do not simultaneously take Urdu case endings when used,
for example, as vocatives or with postpositions; that is, they remain uninflected for case. If Persian
or Arabic loanwords use the Urdu plural endings, however, the usual Urdu inflectional suffixes
will apply in the vocative or oblique case.
ب /masjid mēṁ / in the mosque
وں ب /masjidōṁ mēṁ / in the mosques
ب /masājid mēṁ / in the mosques
* وں ب /*masājidōṁ mēṁ /13 *in the mosques
In contrast, loanwords from other languages, such as English, do not bring their original morphol-
ogy with them, except in the case of words that usually occur in the plural, like matches. In this
case, the plural suffix /-es/ is borrowed with the word, but the word is treated as singular and
inflected like an unmarked feminine noun.
ٹa س/šart ̣/ dress shirt
بa س/šart ̣ēṁ / dress shirts
3.3.1 Persian
Urdu has borrowed many nouns from Persian, as well as many suffixes that can be added to both
indigenous words and loanwords to form nouns.
Persian loanwords or Arabic words that have come into the language via Persian may take Persian
plurals. Commonly used Persian loanwords may also take Urdu plurals, although the reverse is
not true: Urdu words of Indic origin may not take Persian plurals. Using the Persian plural, more
than with other Persian or Arabic elements, nearly always signals a highly formal or literary style.
There are two plural suffixes in Persian:
• ان/-ān/ (with variants گ ن/-gān/, ن/-yān/) originally for nouns that denote animate beings14
58
Nouns
Nouns ending in ہ/-ah/ take the variant گ ن/-gān/; nouns ending in ا/-ā/ take ن/-yān/.
Examples:
/-gī/ is used after words that end in /ah/; /-ī/ is used elsewhere.
Examples:
گ م/garm/ hot → گ/garmī/ heat
16
Sometimes رگa ب/buzurg/ is treated as an unmarked masculine noun; that is, it can be plural without a suffix.
17
The suffix /-cī/ is originally Turkic, but was borrowed into Persian. For the most part, it is not productive in
contemporary Urdu.
59
Nouns
د/dēg/ cauldron, very big pot → د ہ/dēgcah/ largish pot (can be put on stove top)
60
Nouns
• دان/-dān/ ~ دا/-dānī/
Meaning: receptacle for X; the two suffixes are more or less interchangeable, but دان/-dān/ is
masculine and دا/-dānī/ feminine.
The first element is in the oblique case where applicable (i.e., marked masculine nouns).
Examples:
ۓ/cāē/ tea → ۓدا/cāē-dānī/ teapot
/cūhā/ mouse → دان /cūhē-dān/ mouse trap
/namak/ salt → ا /namak-dāni/ salt shaker
• دان/-dān/ Note that this is homophonous with the previous suffix, but has a very different
meaning.
Meaning: knower of X
Examples:
س /syāsat/ politics → دان س /syāsat-dān/ politician
ز ن/zabān/ language → ز ندان/zabān-dān/ knower of languages
Known as ا ہezafeh or izāfat, meaning increase or addition, this enclitic joins two nouns (one of
which can be a pronoun) or a noun and an adjective.19 The first element is always a noun, pronoun,
or verbal participle. The second element modifies or qualifies the first and can be either a noun
or an adjective. In the case of two nouns, the ا ہezafeh conveys a possessive relationship: the
first noun belongs to the second. In theory, the ا ہezafeh can only join Perso-Arabic loanwords,
but in spoken usage it is occasionally used with words of Indic origin.
ا ہEzafe is transcribed as /-e-/ in this grammar. Its representation in Urdu differs according to
the final letter of the first word in the construction:
2. After a long vowel ( اalif or وvāū), ا ہezafeh is spelled with a ءhamzāh, often followed by
a ےbaṛī yē in its independent form.
3. After ہcʰōt ̣ī hē, ہ اezafeh is spelled with a ۂhamzāh over cʰōt ̣ī hē.
4. After ےbaṛī yē, ا ہezafeh can represented by an unseated hamzāh, though it is usually
omitted. The ezafeh is not pronounced.
Increasingly, ا ہezafeh is not pronounced in the spoken language, particularly when the first
element ends in a short vowel + consonant, as in /tālib ilm/ for /tālib-e-ilm/ student.
Examples:
19
See Chapter 6 for information on the izāfat’s function when joining a noun and an adjective.
61
Nouns
Compare this construction with the indigenous Urdu construction that uses the postposition 5
/kā/, /kē/, /kī/:
An ا ہezafeh compound can also form part of a larger compound by being joined in turn with
another word.
This Persian word is used to form compounds in Urdu. The result is an adjective that can also be
used as a noun:
3.3.2 Arabic
Arabic morphology is radically different from the morphology of Urdu (or for that matter, English).
Roots are usually defined by three (occasionally four) consonants to which affixes and vowel
patterns are added to form derived verb and noun forms. These consonants are referred to as
“radicals.” For example, f-c -l is the root for do; it is conventionally used to illustrate patterns of
derived forms, a convention we will follow in the discussion below. The superscript /c / represents
the Arabic consonant ع, a voiced pharyngeal fricative. This sound does not exist in Urdu, but
Perso-Arabic loanwords which have عin their spelling retain it in Urdu. Its pronunciation in
Urdu varies. However, when patterns using the root f-c -l are given, each of the three radicals /f/,
/c /, /l/ stands for any consonant. See below for examples.
As with Persian, heavy use of Arabic loanwords, and more particularly, use of Arabic plurals as
opposed to the Urdu plural formations, connotes a more formal style.
62
Nouns
Adding the suffix 6 /-iyat/ to any noun forms a feminine abstract noun, sometimes with a slang
connotation. This suffix is productive, as can be seen by the last example below, where it has been
added to an English word.22
Examples:
Urdu has a few borrowings--phrases and proper names--that include the Arabic definite article
ال/al/. Learners of Urdu should know one important fact about this Arabic element in Urdu:
although always written the same, ال/al/ is frequently pronounced differently as a result of an
Arabic assimilation rule. Assimilation is a change in pronunciation whereby a consonant or vowel
becomes more similar to a consonant or vowel near it, as in Latin-derived English words like
impossible, imbecile, immobile, where the prefix in- has changed to im- before the labial consonants
p, b, and m.
This assimilation can happen to both the vowel /a/ and the consonant /l/ of the definite article:
20
The citation form for Arabic verbs is the third singular masculine perfect form.
21
The latter is rare in Urdu.
22
A possible analysis of this suffix is that it is actually composed of two suffixes: /-ī/, which forms adjectives from
nouns, plus /-at/, which forms abstract nouns from adjectives, with the meaning the quality denoted by X [= the
adjective]. Thus ا/insān-iy-at/ means the quality of being (a good) human; i.e., humanity.
63
Nouns
1. If the preceding word ends in an /u/, /i/, or /a/, then /a/ assimilates to that vowel. Often
this final vowel is not orthographically explicit; rather, it is the remnant of an Arabic case
ending represented as a diacritic. While this vowel can change in Arabic depending on case
(/u/ for nominative, /i/ for genitive, /a/ for accusative), in Urdu the vowel is usually fixed
depending on the manner in which it was borrowed; most commonly this will be /u/.
2. If the following word begins with a dental or alveolar consonant (،ص،ش،زس،ر،ذ،د،ث، ت/t, s,
d, z, r, š, l, or n/),23 then /l/ assimilates to that consonant and is pronounced as a doubled
consonant.
Examples:
ا ب /šams ud-dīn/ Shams ud-Deen (proper name; lit. sun of the religion)
ا ﷲا /b-ism illāh ir-rahmān ir-rahīm/ in the name of God, Most Compassionate, Most
Merciful (often said when one is beginning something)
ما ا/inzamām ul-haqq/ Inzamam-ul-Haq (famous Pakistani cricketer ; lit. union/associa-
tion of truth)
ا ﷲ/assadullāh/ Assadullah (proper name; lit. Lion of God)
As noted above, if an Urdu noun has an Arabic dual or plural form, it will not take an Urdu case
ending.
Unlike Urdu, Arabic distinguishes three types of number: singular, dual, and plural. Urdu has
borrowed a handful of Arabic nouns which include the Arabic dual suffix ب/-æn/ on them:24
وا/vālid/ father → وا ب/vālidæn/ parents
• ب/-īn/25
This suffix is mostly affixed to participles, as well as some Arabic masculine nouns.
23
These are known as the sun letters in Arabic. Note that while the letter šīn ( )شis generally considered to represent
an alveolar consonant in Arabic, it represents a palatal one in Urdu.
24
This suffix is cognate with (related to) the accusative/genitive dual suffix ِ ب/-ayni/ in Modern Standard Arabic,
but its consonant-final form and usage as a generic dual suffix reveal it to have been borrowed from spoken Arabic,
which dropped the final vowel long ago and extended its use to all cases.
25 َ
Like the Arabic dual suffix, this form is related to an accusative/genitive suffix in Modern Standard Arabic ( ب
/-īna/), but also like that suffix, its usage has been generalized and its final vowel dropped in the spoken dialects of
Arabic, yielding the form that was borrowed into Urdu.
64
Nouns
Examples:
ب /mujāhidīn/ participants in a jihad
بب /mutāsirīn/ affected ones
• ات/-āt/26
This suffix is affixed to nouns of either gender, including verbal nouns. It is also sometimes used
with words of Persian or Indic origin, although the latter only rarely.
Examples:
ت/bāɣāt/ gardens (sg. غ/bāɣ/)
ت ا/imtihānāt/ exams (sg. ا ن/imtihān/)
• ت/-iyāt/
This suffix is used to form the names of branches of learning. It consists of two parts, /-ī/ and
the plural suffix /-āt/.
Examples:
Arabic broken plurals involve altering the vowel pattern of the singular noun. There are too many
patterns to cover all of them here, but the most common are as follows:
65
Nouns
• Pattern: ّ ل/fucc āl/. Note that the second radical is doubled in this pattern.29
Examples:
ّ ر9 /tujjār/ merchants (sg. /tājir/)
ّ ب8 /kuttāb/ writers (sg. 6 5 /kātib/)
• Pattern (vowels only): /a-ā-ī/30
Examples:
وب /tasāvīr/ pictures (sg. ص ب/tasvīr/)
ب5 د/dakākīn/ stores (sg. ن5 د/dukān/)
• Pattern: ا/afic lā/ This pattern is commonly applied to weak roots with a double second radical
that refer to rational beings.
Examples:
ا/asmā/ names31 also: nouns (gram.) (sg. ا/ism/)
66
Nouns
These two prefixes are almost always used with words of Arabic or Persian origin. Exceptions are
listed below.
32
Unlike other Arabic prefixes, this postposition can be attached to adjectives as well as nouns.
33
In Pakistan, the most common meaning is non-Muslim.
67
Pronouns
Chapter 4
Pronouns
4.1 Introduction
Pronouns are the words which stand in for nouns, such as she, he, they, who, this, etc.; the noun
for which a pronoun stands is its referent.
Urdu pronouns have categories for number (singular and plural) in all persons, for degree of
familiarity/respect in the second and third persons, and for distance from the speaker in the third
person. We use the term “proximal” for pronouns that refer to someone relatively near the speaker
and “distal” for those that refer to someone at a distance. There are no gender distinctions in the
pronouns, but this information is often conveyed by the verb. Pronouns take case in the same
manner as nouns: direct case (usually) when they are the subject of a verb and not followed by a
postposition, and oblique case when followed by a postposition, including the “ergative” particle
/nē/.1 , although the pronouns ب/mæṁ /, /ham/, @ /tū/, /tum/, and آپ/āp/ retain
their direct case forms before /nē/.
• Since Urdu verbs also show person, number, gender, and degree of respect, personal pronouns,
especially @ /tū/ and /tum/, can sometimes be omitted. In speech the pronouns are usually
expressed, but they may be dropped in certain contexts, particularly questions, responses to a
question, or in poetry.
• Degrees of familiarity, Second Person: The Urdu second person distinguishes among three
degrees of familiarity, reflecting the relationship between the speaker and the person addressed,
or sometimes, the speaker’s attitude towards that person.
1
See Chapter 5 for discussion of this particle and of the notion of ergativity.
68
Pronouns
With
Direct Oblique /nē/ (Possessive)
1st Person
/ham/ /ham/ /ham nē/
وہ ان ا ں
distal
/voh/ /un/ /unhōṁ nē/
69
Pronouns
1. @ /tū/ (intimate) is reserved for addressing young children in one’s family, God, a beloved,
or animals. It can also be used to express contempt or disapproval, as in an insult or a
reproof to a subordinate. It occurs more often in poetry and film songs than in everyday
speech. @ /tū/ always takes a singular verb.
2. /tum/ (familiar) is used to address equals with whom one is on close terms (and then
only in informal settings; not, for example, in a professional context), servants and others of
lower social status (subordinate co-workers, taxi drivers, servants, waiters, etc.), children
of another family, one’s wife (but generally not one’s husband2 ), and occasionally, one’s
mother.3
/tum/ always takes a plural verb, even if there is only one referent. When applied to
males, it requires a plural noun complement (and adjective, if applicable) in equational
sentences4 , no matter whether there is one male or more than one. When females are
addressed using /tum/, however, the noun complement is singular if only one person is
being addressed and plural otherwise; that is, the complement reflects the actual number
in the case of female referents.
(4.1) ا ّ ڑ
tum accʰ-ē laṛk-ē hō
you good-M.PL.DIR boy-M.PL.DIR be.PRS.2.PL
‘You are a good boy/good boys.’
3. آپ/āp/ (formal) is used to address people of higher status, including family members if
they are older than oneself, skilled people of any rank, or employers; it is used by women to
address their husbands, by people of equal status who are not on intimate or informal terms
with each other or are in formal situations, by young people to address the elderly (includ-
ing servants), and in general, for anyone addressed as /sāhib/ Mr. or بہ/sāhibah/
Mrs./Miss.
آپ/āp/ follows the same rules of agreement as /tum/ above.
2
These uses, however, all vary widely within individual families, and there are some wives who do use /tum/
with their husbands.
3
A few people address their mother as /tum/, but it is not very common except in films. Most use آپ/āp/.
4
An equational sentence is one of the form A is/are B, where B can be either a noun or adjective.
70
Pronouns
• Degrees of familiarity, Third Person: There are also three degrees of familiarity in the third
person, although the distinctions are less commonly made and are not analogous to those of the
second person. But note that when the person referred to is present, it is quite rude to refer
to him or her with a singular form; and since the pronoun /yeh/ does not indicate number,
politness requires a plural adjective form (if applicable) and a plural verb form.
• Third person proximal forms refer to he, she, it, or they here or to the person or thing just
mentioned, and third person distal forms refer to he, she, it, or they over there. The default, and
hence more common, form is وہ/voh/, used when no distinction of distance is intended by the
speaker.
• There are some contexts in which first person plural /ham/ can be used in place of the singular
ب/maiṁ /:
– In colloquial usage, to give a sense of anonymity to the speaker (more commonly with women
speakers)
– In colloquial usage, to connote the speaker’s sense of superiority
– In poetry
• The masculine plural noun گ/lōg/ people, which conveys plurality, can be added after pronouns
for emphasis or disambiguation. Because, for example, the second person plural /tum/, can
refer to one or more people, گ/lōg/ might be used when the speaker wants to make clear that
the referent is plural. The verb would then be masculine plural.
Even though Urdu pronouns, like nouns, generally take postpositions to express grammatical rela-
tions, there is one major exception with the postposition @ /kō/, which is used both with animate
or specified direct objects and with indirect objects.5 Although oblique pronoun + @ /kō/ is ac-
ceptable, it is more common to use an inflected form of the pronoun; that is, pronoun + suffix.
This applies to the personal, demonstrative, interrogative, and relative pronouns, except آپ/āp/.
These forms are sometimes called the special dative forms.A table of these special dative forms of
personal pronouns is below; the special dative forms of demonstrative and relative pronouns are
listed in the tables for those pronominal forms below.
5
See Chapter 5 for more on @ /kō/.
71
Pronouns
The suffixes that are attached to the pronoun are ے/-ē/, ب/-ēm/, or ب/-hēm/, depending on
the form:
Table 4.2 shows the alternate forms for the personal pronouns.
3rd Person
ا/isē/ @اس/is kō/
proximal
Note that the second and third person plural forms can be spelled with either a ( ہcʰot ̣ī hē) or a ھ
(dō cašmī hē), but the former is more common.
(See Table 4.1 in Section 4.2.1) Note that the possessive forms are actually adjectives in form and
agree in gender, number, and case with their nouns.6 They can be used either attributively (as
modifiers: my book) and predicatively (that book is mine). Also note that possessives are of two
types formally: first person and second person intimate and non-honorific possessive pronouns are
single, unique words, while in the second person (honorific) and in the third person, the possession
is expressed with a two-word phrase: oblique pronoun + postposition.
Urdu does not have a complex reflexive pronominal system, but the following forms are used
when a pronoun refers back to the subject of the verb.
6
They are also covered in Chapter 6 on modifiers.
72
Pronouns
To intensify the direct object of the verb, the speaker can place د/xud/ after it, but this still
remains ambiguous, as in the following sentence, where د/xud/ could refer to either ب/mæṁ̇ /
or a ڈا/ḍākt ̣ar/:
4.3.2 6 ا/apnā/
If the subject of a verb possesses the object being talked about, then 6 ا/apnā/ one’s own must
be used, rather than the possessive personal pronoun.Remember that it is an adjective, so it must
agree with the noun it modifies. Note that the second sentence below is ungrammatical, as signaled
by the asterisk.
7
myself, yourself, himself, herself, itself, ourselves, yourselves, themselves
73
Pronouns
(4.8) ؤ6رے ل
4.3.3 د/xud/
This form expresses either an action performed by the subject without help from others or an
action that happens spontaneously:
These phrases may be used with intransitive verbs to express an action that the subject does by
him/her/itself or of his/her/its own accord, or which occurs spontaneously:
74
Pronouns
The demonstratives distinguish number (but only in the oblique case)8 and proximity (see Sec-
tion 4.1 above for a definition of the proximity terms proximal and distal):
Table 4.3 shows demonstrative pronouns.
As in English, Urdu demonstratives can function as adjectives as well as pronouns:
8
The verb will convey whether the subject or direct object (in the case of perfective tenses of transitive verbs) is
singular or plural.
75
Pronouns
Alternate Forms
Direct Oblique
(Special Dative)
بہ ان ا ب ~ا ب
Proximal /ye/ /in/ /inhēṁ /
Plural
these these these.ACC
وہ ان ا ب ~ا ب
Distal /voh/ /un/ /unhēṁ /
those those those.ACC
Relative pronouns are used in relative clauses, which are a type of clause that describes the referent
of a noun or pronoun. For example, in the English sentence Do you see that guy who just came in?
the relative clause is who just came in, and the relative pronoun is who. In Urdu, unlike English, the
relative clause usually comes first in the sentence, and it is generally followed by a parallel clause
beginning with the distal counterpart of the relative pronoun ( وہ/voh/ in the example sentence
below). This parallel clause is known as the correlative clause; for more on relative-correlative
constructions see Chapter 7.
Table 4.4 shows relative pronouns.
Relative adjectives and adverbs are covered in Chapter 6 and Chapter 7 respectively.
76
Pronouns
Alternate Forms
Direct Oblique With /nē/ (Special Dative)
Singular -------------
/jō/ /jis/ /jis nē/
ں ں
Plural
/jō/ /jin/ /jinhōṁ nē/ /jinhōṁ /
Alternatives to + @
Direct Oblique With /nē/ /kō/9
8
Singular /kyā/ /kis/ /kis nē/ /kise/
what?
@ن
/kɔn/ /kise/
who? /kis/ /kis nē/
8 B ب
/kyā/ ----------------
Plural /kin/ /kinheṁ /
what?
@ن B ب
ں
/kɔn/
/kin/ /kinhōṁ nē/ /kinheṁ /
who?
9
These are analogous to the /mujʰē/, etc. forms in Section 4.2.2.
77
Pronouns
(4.18) ر8بہ
Notes
• @ ن/kɔn/ can also sometimes be used as an adjective meaning which (of several)?, but it is
generally accompanied by the particle /sā/ when used this way.See Section 6.6.1 for more
details.
• Because @ ن/kɔn/ does not have distinct singular and plural forms in the direct case, the plural
can be specified by adding گ/lōg/10 or by repeating @ ن/kɔn/. @ ن گ/kɔn lōg/ asks about the
identity of the group as a whole, whereas @ ن@ ن/kɔn kɔn/ is asking which particular individuals
in a group.
The expected answer to the above question would be a type or group; for example, Those who
weren’t invited or Those from Lahore.
10
See Section 4.2.1 above.
78
Pronouns
(4.21) ںآۓ @ ن@ ن
kɔn kɔn tʰē jō yahāṁ ā-ē
who who PST.M.PL REL here come-PP.M.PL
‘Who are the people who came here?’
Here the expected response would be to name individuals; for example, Ahmed, Fatima...
Oblique: /kisī/
@ /kōī/ is used as both a pronoun (someone, anyone) and an adjective (some, any). When used in
a negative sentence, it means no one or no, not any.
79
Pronouns
(4.26) ب8
Like @/kōī/, ھ/kucʰ/ can be used as either a pronoun or an adjective. As a pronoun it means
something in affirmative sentences, and nothing, not anything in negative sentences.
(4.27) ں ھ ب دس ل
As an adjective, ھ/kucʰ/ means some or any and modifies both uncountable and plural count
nouns, but only in the affirmative, not in negative sentences. With count nouns in the negative,
it means not any. Hence the second sentence below is ungrammatical, as signaled by the asterisk.
• Uncountable Nouns
(4.28) ں ھ دل ب
80
Pronouns
(4.29) ب ں ھ دل ب
• Count Nouns
(4.30) ب ں ھ
81
Postpositions
Chapter 5
Postpositions
Postpositions are words that come after nouns, pronouns, or noun phrases and help to convey
the grammatical or semantic relation of that noun to the other words in the sentence. They are
analogous to prepositions in English, except that prepositions come before the noun.
Examples:
We can divide postpositions roughly into two types, with some overlap: those that express primar-
ily grammatical functions, such as possession or object-marking; and those that express spatial or
temporal relationships, such as location or extent.
Although it functions as a postposition and therefore requires that the preceding noun be in the
oblique case, 5 /kā/ is also inflected like an adjective; that is, when used to express a possessive
relationship between two nouns, it must agree in gender, number, and case with the second noun,
which it can be said to qualify. The examples in the sections below illustrate this agreement.1
Table 5.1 shows the postposition 5 /kā/.
1
It is important to remember that 5 /kā/ agrees with the following noun or adjective, not with the one it follows.
82
Postpositions
Direct Oblique
5
Masculine Singular
/kā/ /kē/
Masculine Plural
/kē/ /kē/
5.1.1.1 Possession
~ ~ 5 /kā ~ kē ~ kī/ expresses possession between two nouns or noun phrases. The first noun
(phrase) + 5 /ka/ can be said to form an adjectival phrase modifying or qualifying the following
one:
(5.3) ا
83
Postpositions
(5.7) ں ب ا
nida k-ī māṁ nahīṁ hæ
Neda.DIR POSS-F.SG.DIR mother.F.SG.DIR NEG be.PRS.3.SG
‘Neda has no mother.’
84
Postpositions
(5.9) گ ڑی
šāhid k-ī gāṛī hæ
Shahid POSS-F.SG car be.PRS.3.SG
‘Shahid has a car.’
2. Alienable possession: س /kē pās/ is used to express (temporary) possession or control
of physical or tangible things:
(5.10) سگ ڑی
šāhid k-ē pās gāṛī hæ
Shahid POSS-OBL side car be.PRS.3.SG
‘Shahid does have a car (or Shahid has the car).’
Contrast the sentence above with the previous one. In the previous sentence it is obvious
the car is Shahid’s (inalienable). But in this sentence, the car may or may not be Shahid’s;
the car is simply in Shahid’s possession.
In all of these, 5 /kā/ is inflected to agree with the oblique noun or adjective.
• و ہ /kī vajah sē/ because of, on account of ( و ہ/vajah/ (f.), reason, cause)
85
Postpositions
• ( ) > ذر /kē zarīē (sē)/ by means of ( ذر> ہ/zarīa/ (m.), means)4
• ورب /kē taur par/ as, by way of ( ور/taur/ (m.), manner, way)5
• ب /kē muqābilē mēṁ / in comparison with (ہ /muqābila/ (m.), contrast, comparison,
opposite)
3
The word /sabab/ can also be used without /sē/.
4
This construction has a reversed form, ر> ہ/bazarīa-e-X/, using the Persian preposition /ba/, which is used less
frequently. See Section 5.1.1.2.5 below.
5
This construction also has a reversed form, ور/bataur-e-X/, with the same Persian preposition /ba/. See Sec-
tion 5.1.1.2.5 below.
86
Postpositions
(5.19) ن ۓ5ھد ر ز
maryam zænab k-ē sātʰ dukān jā-ē gī
Maryam.DIR Zainab.OBL POSS-OBL company store go-SBJV.3.SG FUT.F.SG
‘Maryam will go to the store with Zainab.’
87
Postpositions
• وہ /kē ilāvah/ in addition to, besides (وہ /ilāvah ~ alāvah/ (m.), superaddition)
وہ /kē ilāvah/ can also be circumfixed around the noun, as وہ __ /ilāvah X kē/:
(5.25) @ ب ر:را> ہ وہ
rābiah ilāvah arabi k-ē paštō bʰī bōl-t-ī hæ
Rabiah.DIR addition Arabic POSS-OBL Pashto INC speak-IP-F.SG be.PRS.3.SG
‘Rabiah speaks Pashto in addition to Arabic.’
In this construction, 5 /kā/ always takes the form /kē/ in agreement with the oblique adjective
(e.g., /mutāliq/, /mutābiq/, etc.)
(5.26) ب ھ ب ہ ری:
šaīrī k-ē mutāliq mæṁ kucʰ nahīṁ keh sak-t-ī
poetry POSS-OBL connection I.DIR anything NEG say be.able-IP-F.SG
‘I can’t say anything about poetry.’
88
Postpositions
• ب اب /kē barābar/ equal to, similar to, next to ( ب اب/barābar/ even, level, equal6 )
6
From the Persian ب اب/bar-ā-bar/ side by side.
7
Some of the following postpositions are actually cases of 5 /kā/ + oblique noun. However, some oblique nouns
function adverbially and are often perceived as adverbs rather than oblique nouns.
8
Oblique noun.
9
Oblique noun.
89
Postpositions
• a> /kē baɣær/ without (< Arabic ب/bi-/ with + a /ɣayr/ other via Persian)
• ۓ9 /kē bajāē/, /kī bajāē/, instead of (< Per. /ba/ with, by + /jā/ (f.), place)
• و د /kē bāvujūd/ in spite of (< Per. /bā/ with + و د/vujūd/ (m.), existence)
10
Oblique noun. In addition, also shows possession, see list item 2 in Section 5.1.1.1.2.
90
Postpositions
(5.38) و د رآ @
pōlīs-Ø hō-n-ē k-ē bāvujūd cōr-Ø ā-yā
police-DIR be-INF-OBL POSS-OBL in.spite thief-DIR come-PP.M.SG
‘In spite of the police (presence), the thief came.’
• ف ب /kē bar xilāf/ contrary to (< Per. ب/bar/ on, over +ف /xilāf/ (m.), opposition)
• در ن /kē darmiyān/ between, among (< Per. در/dar/ in + ن /miyān/ (m.), the middle)
• ر> ہ/bazarīa-e-X/ by means of X (< Per. /ba/ with, by + ذر> ہ/zarīa/ (m.), means)11
• ور/bataur-e-X/ as, by way of X (< Per. /ba/ with + ور/taur/ (m.), manner)12
91
Postpositions
Some verbs use 5 /kā/ to mark their objects. See Section 5.3 below for examples.
5.1.2 @ /kō/
1. The speaker wishes to show that the direct object is unspecified (by putting the direct object
in the direct case), or
2. the verb is one that takes a different postposition13 to mark what would be a direct object
in English.
Inanimate direct objects are not usually marked with a postposition, but may be under certain
circumstances; i.e., to indicate strong, directed action toward the direct object, or to stress the
affectedness of the direct object.
13
5 /kā/, /se/, ب/par/
92
Postpositions
(5.47) ب د 8 @ ب
In many Urdu sentences, the semantic subject of the verb takes @ /kō/, while the verb agrees with
what would be the object in the English gloss. These are referred to as indirect or impersonal con-
structions or dative subjects, and the noun is called the logical subject, or sometimes the experiencer
subject. It is regarded as the subject even though it does not agree with the verb because it con-
trols the reflexive possessive pronoun 6 ا/apnā/14 , can be the subject of a conjunctive participial
clause15 , and usually comes first in the sentence.
These constructions generally have to do with a subjective experience of the subject; for example,
liking, remembering, feeling. The verbs commonly used in indirect constructions are /hōnā/ to
be, /ānā/ to come, بڑ/paṛnā/ to fall, and /lagnā/ to be applied, attached.
14
See Section 4.3.2.
15
See Chapter 8.
93
Postpositions
(5.50) ری ا @د ہ
(5.51) ا@ @ ٹ
@ /kō/ is used to express the possession of intangible or abstract things like time, illness, etc.
5.1.3 /nē/
/nē/ was historically used only in the perfective tenses, where it obligatorily marks the logical
16
See Section 5.1.1.1.2 above.
94
Postpositions
subject of transitive verbs, verbs that take an object. In these constructions, the subject is in the
oblique case, and the direct object, when not marked with @ /kō/, is in the direct case. The verb
agrees with the direct object. /nē/ is not used if the verb is intransitive, and in non-perfective
cases of transitive verbs.
Some grammarians call /nē/ an ergative marker.17
(5.55) ا ن
It has been observed that the use of /nē/, particularly in Pakistani Urdu, appears to be evolv-
ing. It can be used with certain intransitive verbs, for example, to indicate purposefulness, or
volitionality, on the part of the subject.18
(5.57) رام
versus
(5.58) رام
Which intransitive verbs can take /nē/ varies by dialect, with Delhi Urdu’s list considerably
shorter than, say Pakistani Urdu’s.
Pakistani Urdu has developed another use of /nē/ in infinitive constructions, where it contrasts
with @ /kō/. If @ /kō/ follows the subject, the implication is that the action takes place as a result
17
See Section 8.5 for more discussion of /nē/ as an ergative marker.
18
This observation and the following two sentences are from Bashir (1999).
95
Postpositions
of external circumstances; that is, the subject is obliged or compelled. But the use of /nē/
after the subject implies either volition on the part of the subject or neutrality with reference to
the subject’s intent. (See Section 8.6.3.2.1.)
5.2.1 /sē/
In a general sense, the semantics of /sē/ may be viewed as having to do with the origin of an
action: its place of origin; its time of origin (time from which); or its instrument, cause, or agent.
“Ablative” refers to movement away from something; in the case of Urdu, this movement can be
in either time or space, so /sē/ in its ablative sense usually means from (spatial) or for, since
(temporal).
96
Postpositions
• Agentive: /sē/ also marks the logical subject of verbs that convey involuntary action, which
Urdu expresses through intransitive verbs, verbs that do not take an object:19
• With passive verbs: /sē/ is used in two different senses with passive verbs.
1. Normal passive
(5.64) گ دو
kʰān-ā jādū sē pakā-yā ga-yā
food-M.SG.DIR magic from cook-PP.M.SG go-PP.M.SG
‘The food was cooked by magic.’
2. To express incapacity
(5.65) ا وا تا ظ ب ھ
mujʰ sē æsē vāhiyāt alfāz nahīṁ kah-ē jā-t-ē
I from such vulgar word.PL NEG say come-IP-M.PL
‘I can’t say such vulgar words.’
5.2.1.3 Comparison
97
Postpositions
Added to certain types of nouns, /sē/ can form an adverbial phrase. See Section 7.1.3.1 for
examples.
Some verbs use /sē/ to mark the words they govern (what would be their direct objects in
English). See Section 5.3 below for examples.
As in the English prepositional collocations out of, up from, over towards, Urdu postpositions can
combine to form more complex sequences. Such sequences with /sē/ always start with a
spatial-temporal postposition and end with /sē/, which expresses movement.
(5.67) گگ ر ر
5.2.2 /tak/
98
Postpositions
Note that /tak/ can sometimes be used as an emphatic particle, meaning even, as well as a
postposition. When used this way, /tak/ does not trigger the oblique case in nouns.
5.2.3 ب/mēṁ/
ب/mēṁ / has the basic meaning of in, and takes on a variety of functions, listed below.
5.2.3.1 Locative
The locative use of ب/mēṁ / can apply to location in either space or time.
99
Postpositions
This usage often occurs with with the verb /lagnā/ to be applied.
100
Postpositions
The locative sense of ب/par/ differs from that of ب/mēṁ /: ب/par/ expresses location on a
surface, next to something, or right at or immediately after a point in time.
sāt baj kar cʰe minat ̣ par alī gʰar pohoṁ c-ā
seven strike CP six minute on Ali house reach-PP.3.M.SG
‘Ali reached home at six minutes past seven (o’clock).’
20
This variant occurs in both the colloquial language and poetic usage.
101
Postpositions
(This usage is idomatic; in this sense ب/par/ and بہ/peh/ do not work for most other places; e.g.,
home, school, etc.)
Some verbs use ب/par/ to mark the nouns they govern. See Section 5.3 below for examples.
As in English, with prepositions,21 many Urdu verbs take particular postpositions with their ob-
jects. Which postposition to use with a particular verb is called a lexical property of that verb; that
is, it is intrinsic to the verb. It must therefore be memorized by speakers.
The postpositions that occur with verbs are:
21
Compare English wait for, look at, etc.
102
Postpositions
5.4 Prepositions
Although postpositions predominate, Urdu has borrowed a few prepositions from both Persian
and Arabic. Most are listed in Urdu dictionaries.
Urdu has borrowed several Persian prepositions. They are usually written as part of the noun they
go with, with the exceptions of از/az/, ب/bar/, /bē/, and /tā/:
103
Postpositions
• /bē/ without 23
ن /bē jān/ lifeless ( ن/jān/ life, soul)
ہ /bē qāida/ abnormal (ہ /qāida/ rule, base)
ر /bē rang/ achromatic, colorless ( ر/rang/ color)
• در/dar/ in
در/dar haqīqat/ actually, in truth ( /haqīqat/ truth)
• /tā/ up to
دن رات/din tā rāt/ from morning to night
Four of these-- ب/ba/, /bā/, ب/bar/, and در/dar/--combine with nouns to form compound
postpositions. See Section 5.1.1.2.5 above.
• /fī/ per
Examples:
ی /fī sadī/ per cent
ا ل/fī-l-hāl/ for now
ّ
ا نﷲ/fī imān allāh/ in the protection of God (said as a farewell)
• ب/bi/ with, in
Examples:
ّ
ﷲ/bi-sm allāh/ in the name of God
/bi-l-kul/ completely
رض /bi-l-farz/ for example
• ل/li/ for, to
Example:
ّہ ا/al-hamdu li-llāh/ praise be to God
23
Compare the English prefixes a-, an-, un-, and the suffix -less.
104
Adjectives
Chapter 6
Adjectives
6.1 Inflection
Urdu adjectives fall into two broad classes: marked and unmarked. As with the noun classes
discussed in Chapter 3, unmarked adjectives have a bare stem in the citation form, while marked
adjectives employ a masculine direct singular suffix.
Marked adjectives agree with the nouns they modify in gender, number, and case, through a small
set of inflectional suffixes. Unmarked adjectives, unlike unmarked nouns, have only a single,
invariant form.
Since adjectives are inflected for the same features that nouns are inflected for, the formal gram-
mar’s listing of morphosyntactic features for nouns (given in Chapter 3) suffices for adjectives as
well.
The normal inflectional paradigm has three endings: ا/-ā/ (masculine singular direct, the citation
form); ے/-ē/ (masculine singular and plural oblique); and ی/-ī/ (all feminine forms).
Table 6.1 shows the declension of the regular inflected (“marked”) adjective و/cʰōt ̣ā/ small.
The formal grammar for inflected (“marked”) adjectives is given below.
105
Adjectives
<!--Four of these are "real" suffixes, while the fifth is a dummy (null)
suffix that attaches to "unmarked adjectives." This makes the
analysis of unmarked adjectives parallel to that of unmarked
nouns, although unmarked nouns do take non-null suffixes in some
paradigm cells.-->
<Mo:refInflectionalAffix idref="afAdjMascSgDirect"/>
<Mo:refInflectionalAffix idref="afAdjMascSgOblVoc"/>
<Mo:refInflectionalAffix idref="afAdjMascPl"/>
<Mo:refInflectionalAffix idref="afAdjFem"/>
<Mo:refInflectionalAffix idref="afUnmarkedAdjSuffix"/>
</Mo:InflAffixSlot>
The suffixes marking gender, number and case on adjectives are defined here.
<!--Masculine-->
<Mo:InflectionalAffix gloss='-MascSgDirect' id='afAdjMascSgDirect'>
106
Adjectives
<Mo:AffixAllomorph>
<!-- -ā , used on most marked adjectives-->
<Ph:Form spelling='+'ا/>
<Mo:inflectionClasses>
<Mo:refInflClass idref='classAdjPlain'/>
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Mo:AffixAllomorph>
<!-- -āṁ , used on nasal declension class adjectives-->
<Ph:Form spelling='+'اں/>
<Mo:inflectionClasses>
<Mo:refInflClass idref='classAdjNasal'/>
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvMasculine'/>
<Fs:SymbolicFeatureValue defn='fvSingular'/>
<Fs:SymbolicFeatureValue defn='fvDirect'/>
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
107
Adjectives
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
<!--Feminine-->
<Mo:InflectionalAffix gloss='-Fem' id='afAdjFem'>
<!--Syncretic in all three cases, singular and plural -->
<Mo:AffixAllomorph>
<!-- -ī , used on most marked adjectives-->
<Ph:Form spelling='+'ی/>
<Mo:inflectionClasses>
<Mo:refInflClass idref='classAdjPlain'/>
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Mo:AffixAllomorph>
<!-- -īṁ , used on nasal declension class adjectives-->
<Ph:Form spelling='+ 'ب/>
<Mo:inflectionClasses>
<Mo:refInflClass idref='classAdjNasal'/>
108
Adjectives
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvFeminine'/>
<!--Since this is used for all three cases and for both singular
and plural, we don't mention the case or number features here.
-->
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
The ordinal numbers and the words for right and left nasalize these inflectional endings: اں/-āṁ /
(masculine direct singular, the citation form); ب/-ēṁ / (masculine oblique and plural); and ب
/-īṁ / (all feminine forms).
Note that in the citation form, the long ا/ā/ vowels of the stem and the masculine singular di-
rect ending must be separated by a glide, ی/y/. This alternation also occurs in the perfective
participles; see Section 8.1.4.
Table 6.2 shows suffix ں/-ṁ / marking gender, number, and case on adjectives.
Adjectives modifying noun sequences generally agree with the nearest noun in number and gender.
109
Adjectives
دا ں
دا ب دا ب دا ب
Masculine /dāyāṁ / /dāeṁ / /dāeṁ / /dāeṁ /
right
ں
ب ب ب
/bāyāṁ / /bāeṁ / /bāeṁ / /bāeṁ /
left
(Observe that the oblique plural suffix is added to cardinal numbers to indicate totality.)
Unmarked adjectives have the same form in all cases, numbers, and genders, and are likely to be
non-Indic borrowings. Since new adjectives are still being borrowed, particularly from English,
the class of unmarked adjectives is growing.
Some nouns and adjectives, many of them Persian past participles, end in a ‘silent’ cʰot ̣ī hē. (See dis-
cussion of nouns in this form in Section 3.1.1.1.1, and the note on transcription in Section 2.2.2.5).
110
Adjectives
Among the nouns, this ending marks the citation form of a marked paradigm. Adjectives with this
ending, however, are generally unmarked.
Example:
ا † ہرa /mērā pasandīdah rang/ my favorite color
ہ ی † ہa /mērī pasandīdah xālah/ my favorite maternal aunt
However, a few, like زہ/tāzāh/ fresh, may take both marked and unmarked endings. Whether an
adjective in this class may take the marked inflection is difficult to predict.
Example:
Long /-ī/ or /-ā/, spelled with alif, are common endings for adjectives borrowed from Arabic:
Example:
ر /ālā mēyār/ (m.), a high standard
Also of Persian origin are a number of suffixes which form adjectives from nouns.
6.1.2.3.1 In ی/-ī/
This suffix, identical to that found in some Arabic borrowings, derives descriptive adjectives from
nouns. Its meaning corresponds roughly with the English suffixes /-al/ and /-ish/: of, pertaining
to. It may be used with both native Urdu and borrowed Persian or Arabic vocabulary items.
د/dēsī/ indigenous (from د/dēs/ country)
111
Adjectives
This is also the suffix seen in such adjectives of origin as † /pākistānī/ Pakistani, /panjābī/
Panjabi, /bangālī/ Bengali, etc. (See Section 3.2.3.4.)
The ی/-ī/ suffix may also derive adjectives from other parts of speech, or may create secondary
adjectives from existing adjectives, as ا رو/andarūnī/ interior (adj.) from ا رون/andarūn/ inner
(adj.).
6.1.2.3.2 In ابہ/-ānah/
This suffix derives adjectives of quality; compare the English suffix /-ly/.
Example:
دو بہ/dōstānah/ friendly (from دوس/dōst/ friend)
ﻻبہ/sālānah/ annual (from ل/sāl/ year)
Adjectives in ابہ/-ānah/ cannot be used to describe humans, though they are usually formed from
nouns referring to people or types of people.
These suffixes correspond roughly to English /-ful/; they create adjectives of quality from abstract
nouns. Neither is productive--that is, new words are not formed with these suffixes.
In general, ک/-nāk/ refers to the cause of the described result (generally something harmful),
and ب/-gīn/ to its sufferer: کa س/šarmnāk/ means causing shame, while بa س/šarmgīn/ bashful
could also be defined as subject to shame; affected by shame.
The suffixes دار/-dār/, اوار/-āvar/ or وار/-vār/, ور/-var/, ر/-yār/ or ب/-yar/, and /-mand/ also
form denominals, all with the general meaning of possessing X, characterized by X.
Examples:
ار ا/īmānda̅r/ faithful, trustworthy (from ا ن/īmān/ faith, integrity)
وارa /hunarvār/ skillful (from a /hunar/ skill)
112
Adjectives
Unlike with ک/-nāk/ and ب/-gīn/, there is no predictable difference in meaning among these
suffixes; where a stem may form adjectives with two or more of them, the precise meanings of the
derived words must be learned separately.
/-mand/ and ر/-yār/ are not productive, but دار/-dār/ is used in new compounds with all
classes of Urdu words--Indic, Perso-Arabic, or recent borrowings from other languages. It can also
produce nouns, as ز ار/zamīndār/ landowner, from ز ب/zamīn/ land.
وار/-vār/ can also be a nominal ending, as in وار6 /pædavār/ production, income.
Some words with the ور/-var/ suffix have been reanalyzed as nouns such as @ ر/jānvar/ animal
(lit. possessing life)
The prefix /nā-/ derives umarked adjectives from nouns and bare verb stems:
/nā-haqq/ unjust (< /haqq/, right)
Some otherwise unmarked adjectives may only be used predicatively in their citation forms: that
is, they may only be used as the complement of a state-of-being verb, and not in prenominal
position. Note that the second sentence below is ungrammatical, as signaled by the asterisk.
113
Adjectives
The suffix ہ/-āh/ (spelled in cʰot ̣ī hē, homophonous with the suffix discussed in Section 6.1.2.1)
makes such predicative-only adjectives attributive, allowing them to be used to directly modify
nouns.
Adjectives of this type are Arabic, and tend to be derived from participles, but there is no rule to
predict which Arabic adjectives must be modified in this way to be used predicatively.
(6.5) ی وبہa
6.3 Numbers
In Urdu orthography, numerals are written from left to right, as in the Indic scripts.
114
Adjectives
6.3.1 Cardinals
The cardinal numbers are unmarked, though they are considered grammatically masculine.
The forms of the cardinals from one to one hundred are not totally predictable by rule-- their
compounded forms are idiosyncratic and must be learned individually.
Numbers above one hundred are formed using the following units:
ارa /hazār/ one thousand (1,000)
سو رہ9 ارa دو/dō hazār pāṁ c sɔ bāra/ two thousand, five hundred twelve (2,512)
These units have gender, and the numerals that count them agree with them in gender, but they
are otherwise uninflected.
The oblique plural, added to cardinal nouns, gives the meaning of totality: /cārōṁ / all four;
/donōṁ / both, etc.
6.3.2 Ordinals
As mentioned above in Section 6.1.1.2, most of the ordinal numbers follow the paradigm for the
words for right and left. They are formed with the suffix واں/-vāṁ /: e.g., 9 /pāṁ c/ five, and
اں9 /pāṁ cvāṁ , ب9 /pāṁ cvēṁ /, ب9 /pāṁ cvīṁ / fifth.
Five of the ordinals are instead formed on the standard /ā, ē, ī / paradigm, as in Section 6.1.1.1:
/pahlā/ first; اa دوس/dūsrā/ second; اa /tīsrā/ third; /cɔtʰā/ fourth, and /cʰat ̣ā/ sixth.
The word @ اں/navāṁ / ninth, is also sometimes considered irregular, on account of an alteration in
its stem: the vowel in /nɔ/ nine, becomes a short /a/ when the ordinal suffix is added. However,
it inflects like the other ordinals: @ ب، @ ب، @ اں/navāṁ , navēṁ , navīṁ /.
In the ordinals 11-17, the final silent cʰot ̣ī hē of the cardinals appears as a do cašmi he. This
aspiration may be pronounced as an independent short syllable: so, گ رہ/gyārah/ eleven, yields
گ ر اں/gyārʰvāṁ / eleventh.
115
Adjectives
6.3.3 Fractions
The fractions @ /pɔnā/ three quarters (of ), and آد/ādʰā/ half (of) are inflected. All other words-
-nouns and adjectives--for fractions and fractional quantities are indeclinable.
Fractional quantities up to and including one and a half are grammatically singular, though there
is much variation on this point in the spoken language.
6.3.4 Iteration
There is no special morphology for once, twice, etc. These meanings are expressed with the number
and either the word ر/bār/ time, occasion or د ہ/dafā/ occasion, moment: ا د ہ/ek dafā/ one
time; دو ر/do bār/ two times, and so forth.
The adjectives 6 ا/itnā/ this much, and ا/æsā/ like this, belong to a class of four-part word sets:
near and far demonstratives, interrogatives, and relatives. These functions are distinguished by
initial sound: ی/y-/ (or /i-/, /æ-/, or /a-/) for the near words, و/v-/ or /u-/ for the far words,
/k-/ for the interrogatives, and ج/j-/ for the relatives. Compare this to the symmetry in English
here/there/where, hither/thither/whither, hence/thence/whence.
Table 6.3 shows the y-v-k-j class of demonstrative, interrogative, and relative adjectives 6 ا/itnā/
this much, and ا/æsā/ like this. 6 ا/itnā/ and ا/æsā/ are regularly inflected adjectives,
6ا 6ا
/itnā/ /utnā/
this much; these that much; that many; /kitnā/ /jitnā/
how much? as much
many; so many so many
ا و
/æsā/ /væsā/1
such (a); like this; of such (a); like that; of /kæsā/ /jæsā/
how? such as
this kind that kind
following the patterns in Section 6.1.1.1 and Section 6.1.1.2. They may also be used as nouns,
and as adverbs when inflected in the masculine singular oblique (see Section 7.1.2.1).
The form /tæsā/, part of the near-extinct fifth class of correlatives, survives in a handful of frozen
phrases.
1
و/væsā/ is used primarily in spoken Urdu. In written Urdu, when the meaning is emphatic (such or such a), ا
/æsā/ is preferred, even for remote referents.
116
Adjectives
6 ا/itnā/ and 6 ا/utnā/ tend not to occur before singular nouns, except for mass nouns--nouns that
do not form plurals, like water, snow, or milk.
The larger class of y-v-k-j word sets includes personal pronouns and adverbs of place, time, and
manner; see Section 4.4 for the demonstrative pronouns.
The interrogative pronouns 8 /kyā/ which, and @ ن/kɔn/ who, may both be used as adjectives; see
Section 4.6.
6.5 Comparison
Comparative and superlative forms may function comparatively, or they may be used as intensives,
with the meaning very X.
Some Persian-derived adjectives form the comparative with the suffix ب/-tar/, and the superlative
with ب ب/-tarīn/
Examples:
بہ/beh-/ good (appears only as a prefix)
a /behtar/ better
بa /behtarīn/ best
These suffixes may also be written as freestanding words, with no change in meaning; this is a
more formal usage than the phrasal comparatives below in Section 6.5.2, and is rarely used, ( )س
/(sab) sē/ being the preferred construction.
/jadīd/ new
ب /jadīdtar/ newer
بب /jadīdtarīn/ newest
ص رت /xūbsūrat/ pretty
ص رتب /xūbsūrattar/ prettier
ص رتب ب /xūbsūrattarīn/ prettiest
These comparatives are absolute, rather than relative: بa /behtarīn/ means not the best of all
(of a set), but superlative, excellent.
Some Persian adjectives, such as بہ/beh-/ good and /bad/ bad, form comparatives only with
ب ب، ب/tar, tarīn/. Other Persian adjectives can take either phrasal comparatives or ب ب، ب/tar,
117
Adjectives
tarīn/. For these adjectives which can take both (e.g., ص رت/xūbsūrat/ pretty), the ب ب، ب/tar,
tarīn/ form is the more elevated; however, these elevated forms are most commonly used for
ironic or humorous effect, and consequently are usually very colloquial.
Indic adjectives rarely take ب ب، ب/tar, tarīn/, and generally are only for ironic effect. However,
the suffix appears more commonly on English borrowings, e.g., cheap-tarīn, stupid-tarīn.
In colloquial use, the superlative ب ب/tarīn/ is more common than the comparative ب/tar/.
6.5.2 Phrasal
6.5.2.1.1 Comparatives
The comparative may be formed phrasally with the word /sē/, a postposition meaning from--or,
in this instance, than--following the modified noun.
ص رت ھ/mujʰ sē xūbsūrat/ prettier than me
6.5.2.1.2 Superlatives
The superlative is formed with س/sab sē/ of all. An absolute superlative--the most X of all--
may be constructed, of the form Adj /se/ Adj, with both adjectives in their citation form.
ص رت س/sab sē xūbsūrat/ prettiest of all
ص رت ص رت /xūbsūrat sē xūbsūrat/ prettiest of the pretty
ز دہ/zyāda/ more, and ز دہب/zyādatar/ most, may also be used to form phrasal comparatives and
superlatives. These words precede the adjectives they modify.
Both words are also used adverbially.
، ، /sā, sē, sī/, following a noun or an oblique personal pronoun, forms an adjectival phrase
with the meaning like X. /sā/ agrees with the noun this adjectival phrase modifies in case,
number, and gender.
Table 6.4 shows the declension of /sā/ in phrasal adjectives.
118
Adjectives
Direct Oblique
Masculine Singular
/sā/ /sē/
Masculine Plural
/sē/ /sē/
(6.7) @ ن گ گ
When @ ن/kɔn/ is followed by a form of /sā/ (/sā ~ sē ~ sī/), only the particle inflects; oblique
forms of @ ن/kɔn/ are not used, even where they would otherwise appear. /kɔn sā/ has the
meaning Which (of several)? and is used most commonly of inanimate nouns.
/sā/ following an adjective corresponds in meaning to English /-ish/, lessening the force of the
adjective.
a ّ ا/accʰī xabar/ good news (a /xabar/ is feminine.)
a ّ ا/accʰī sī xabar/ goodish news, rather good news
و ڈبّہ/cʰōt ̣ā dabbah/ a small box
119
Adjectives
وا، وا، واﻻ/vālā, vāle, vālī/ is an adjectival formant, inflected like � /kā/. It forms phrases
meaning with (the) X when it follows an oblique noun, or engaged in X-ing when it follows an
oblique infinitive verb. (See Section 8.6.4 for more on the use of واﻻ/vālā/ with infinitives.) It
can also be added to adjectives in colloquial speech.
واﻻ/vālā/ agrees with the noun modified by the adjectival phrase.
• Like most Urdu adjectives, واﻻ/vālā/ adjectives can also function as nouns, to denote someone
who does, sells, possesses, or otherwise has to do with something.
• Though, as marked adjectives, واﻻ/vālā/ adjectives may appear in masculine or feminine forms,
وا/-vālī/ forms do not occur in the feminine plural; instead, an inflected form of ڑ/laṛkī/ or
something similar is added.
• Nominal examples:
• Adjectival examples
زارواﻻ/bāzār-vālā kʰānā/ food from the market
ب8 ا ّ وا/acchī-vālī kitāb/ the good book3
2
@ /bolnē/ is the infinitive of the verb to speak
3
When واﻻ/vālā/ is added to an adjective, the result is not a change in meaning of the adjective, but a more colloquial
feel and a stressing of which object is meant: the good book, (as opposed to the not-so-good one).
120
Miscellaneous
Chapter 7
Miscellaneous
7.1 Adverbs
Unlike nouns and verbs, adverbs do not change form with the structure of the sentence.1 Ac-
cordingly, most of the morphological information that we can learn from an adverb’s form is
derivational, showing from what elements the word has been formed.
This means that most adverbs may simply be looked up in a dictionary in the form in which they
appear.
Many Urdu adjectives and adverbs belong to a class of four-part word sets: near and far (or prox-
imal and distal) demonstratives, interrogatives, and relatives. These functions are distinguished
by initial sound: /y-/ (or /i-/, /æ-/, or /a-/) for the proximal words, /v-/ or /u-/ for the distal
words, /k-/ for the interrogatives, and /j-/ for the relatives. Compare this to the symmetry in En-
glish here/there/where, hither/thither/whither, hence/thence/whence. (See Section 4.4, Section 4.5,
and Section 4.6 on demonstrative, interrogative, and relative pronouns, and Section 6.4 on the
related adjectives.)
Table 7.1 shows the y-v-k-j class of demonstrative, interrogative, and relative adverbs.
The members of the set ، ، اس، اس/is, us, kis, jis/ are all oblique forms of the demonstra-
tive, interrogative, or relative pronouns; while they are not adverbs themselves, they are used in
adverbial phrases of place, time, or manner (see Section 4.4), and are included in this table for
completeness.
Properly, the words 6 /tab/ then, and @ ں/tūm/̇ thus, so, belong to an archaic class (of correlatives,
in /t-/), but as their meanings and functions fit the u/v class, they are considered a part of it.
The members of the set ، ، و، ا/æsē, væsē, kæsē, jæsē/ are all oblique forms of adjectives
(see Section 6.4); their adverbial use is restricted to the masculine oblique singular.
1
However, adverbs may agree in two circumstances. Adverbs of quantity may agree with the nouns to which they
relate. And adjectives used adverbially, such as بڑی/baṛī/ great(ly), very agree with the modified element.
121
Miscellaneous
اس اس
/is/ /us/ /kis/ /jis/
in this in that in which? in such
اب 6 8
/ab/ (/tab/) /kab/ /jab/
now (then) when? when
ں
ں
@ں ں
/jūṁ /
(/tyūṁ /) as (rare; more
/yūṁ / (thus, so (extremely /kyoṁ /
thus why? common when
rare))
reduplicated)
ا و
/æsē/ /væsē/ /kæsē/ /jæsē/
like this like that how? such as
6ا 6ا
/itnē/ (/utnē/) /kitnē/ (/jitnē/)
this much, such (that much) how much, how ! (as much)
122
Miscellaneous
ا/itnē/ and /kitnē/ are also adverbial uses of adjectives; the other members of this set are
not used as adverbs.
Words of the k class, the interrogatives, are used in questions. Words of the j class, the relatives,
are used in relative clauses--that is, in clauses which refer back to some earlier noun and describe
it (see also Section 4.5).
The interrogative class are also commonly used in rhetorical questions, to ironically undercut or
outright deny the sense of the sentence.
This set of words is comprised of the oblique forms of four pronouns: بہ/yeh/ this, وہ/vɔ/ that,
8/kyā/ what?, and /jo/ which. These words combine with the nouns و/vaqt/ time, ہ/jagah/
place, رف/taraf/ direction, and رح/tarah/ way to form adverbial phrases. The nouns in these
phrasal adverbs must be in the oblique case.
Some of the adverbs of time, اب/ab/ and 8 /kab/ (but not 6 /tab/ or /jab/ when), may
be compounded with the emphatic particle /hī/:
ا/abʰī/ right now
123
Miscellaneous
Nouns, too, may be used adverbially when inflected in the oblique singular. This inflection is
only visible on marked masculine nouns, though feminine and unmarked nouns used in this way
are considered oblique. Any demonstrative used with adverbial nouns must also be oblique. This
usage is especially common with adverbs of place and time.
Adverbial phrases of manner may be formed from nouns with the postposition /sē/ from.
/xāmōšī sē/ with quiet, quietly
زور/zor sē/ with force, forcefully, loudly
/xušī sē/ with happiness, gladly
ً
7.1.4 Arabic Adverbs in ا/-an/
ً
The Arabic adverbial ending ً ا /-an/, which is seen on some borrowed words in Urdu, is spelled
with the symbol tanvīn ()ا, written over alif.
ً
را/fauran/ immediately
ً
را/majbūran/ compulsorily, under duress
ً
/maslan/ for example
The meanings of these adverbs and the nouns from which they ً are derived may have diverged,
and in some cases the senses must be learned separately : e.g., 6 ر/taqrīban/ approximately, from
6 ر/taqrīb/ ceremony (lit. bringing near).
124
Miscellaneous
7.2 Particles
An emphatic particle emphasizes the word or phrase which precedes it. Such particles may be
free or bound-- that is, they may appear as freestanding words, or as suffixes on other modifiers.
The particle @ /tō/ follows the noun or verb in main clauses, for contrastive emphasis: The em-
phasized word is contrasted with some other possibility, often left unstated.
yahāṁ tō kōī hō gā
here EMPH someone be.SBJV.3.SG FUT.M.SG
‘There must be someone HERE.’
yahāṁ kōī tō hō gā
here someone EMPH be.SBJV.3.SG FUT.M.SG
‘There must be SOMEONE here.’
@ /tō/ marks the result clause in a conditional sentence; see discussion in Section 8.3.2.2.2. Unlike
اگ/agar/, which introduces the condition clause, it may only rarely be omitted; accordingly, many
conditional sentences are marked only by @ /tō/.
Condition clauses consisting of commandsCondition clauses may be linked to their result clauses
by the compound conjunction /nahīṁ tō/ if not... then, otherwise:
125
Miscellaneous
In relative sentences beginning with /jab/, @ /tō/ may mark the correlative clause.
The particle /hī/ emphasizes the word which precedes it, and excludes other possibilities (which
may be left unstated). Compare the shifting emphasis in the following two examples:
(7.10) †نب ا
/hī/ may separate a noun or pronoun from its postposition, or it may follow the postposition;
either is correct.
126
Miscellaneous
(7.11) 6د @
(7.12) 6@ د
The exception is the postposition /nē/ with the pronouns ب/mæṁ / and @ /tū/; these pro-
noun+postposition combinations may not be broken up by /hī/. Note that the second sentence
below is ungrammatical, as signaled by the asterisk.
With an oblique imperfect participle, or with the oblique relative /jæsē/, /hī/ denotes
immediate action: as soon as.
(7.15) 6ب ڈالد 6aۓس بف
/hī/ can be inserted in-between two repeated nouns to give the meaning of nothing but x.
127
Miscellaneous
(7.16)
The particle /hī/ may appear suffixed to the oblique forms of all pronouns except the following:
آپ/āp/; the nominative forms of وہ/voh/ and بہ/ye/; and the adverbs of time ( اب/ab/, 6 /tab/,
8 /kab/, /jab/) and place ( ں/yahāṁ /, و ں/vahāṁ /, ں/kahāṁ /, ں/jahāṁ /). In these
bound forms, it is written together with the pronoun it attaches to.
In some such compounds, /hī/ appears in the forms ی/-ī/ and ب/-īṁ /.
Table 7.2 shows bound forms of /hī/ with personal pronouns. Notice that in the bound forms,
the dō-cašmī-heh is used rather than the cʰot ̣ī heh.
See Section 7.1.1.3 above for forms of time and place adverbs with suffixed /hī/.
The suffixed word ب/kahīṁ / somewhere, is also delexicalized and used with or in place of اگ
/agar/. In this usage, it denotes improbability, or, with بہ/na/ apprehension.
The particle /bʰī/ marks the word which precedes it as one of several possibilities: in this
usage, it may be translated with too, also.
128
Miscellaneous
Direct Oblique
Singular
ب
1st person /mæṁ hī/ /mujʰī/
I me
@ 9
2nd person
(intimate) /tū hī/ /tujʰī/
you you
و ا
3rd person
proximal /vahī/ /usī/
he, she, it him, her, it
و ا
distal /yahī/ /isī/
he, she, it him, her, it
Plural
ب
1st person /hamīṁ /
/ham hī/
we us
ب
2nd person
informal /tum hī/ /tumhīṁ /
you you
آپ آپ
formal
/āp hī/ /āp hī/
you you
و ا ب
3rd person
proximal /vahī/ /inhīṁ /
they they
ا ب
distal /yahī/ /unhiṁ /
they they
129
Miscellaneous
Especially when following a noun, /bʰī/ may also be translated as even; following a verb, it
may be translated as just.
When following the indefinite pronouns @ /kōī/, and ھ/kucʰ/, /bʰī/ is translated at all.
/bʰī/ may introduce the second clause of a parallel construction, where the first clause is
introduced by ب /hī nahīṁ /. The entire construction may be translated with not only...but
also.
130
Miscellaneous
/bʰī/ forms phrases with ر /phir/ then; /tō/, the contrastive particle (see Section 7.1.1); اور
/ɔr/ and; and ب/par/ on.
ر /pʰir bʰī/ even so (coordinating conjunction)
@ /tō bʰī/ (also /tab bʰī/) still, nevertheless (coordinating conjunction)
اور/ɔr bʰī/ even more (qualifier; may precede a noun or an adjective)
ب/par bʰī/ despite (coordinating conjunction; follows an oblique infinitive)
/bʰī/ may follow the relatives /jō/, ں /jahāṁ /, /jab/, and /jæsā/. To these words,
it adds the meaning -ever.
131
Miscellaneous
7.2.2 Conjunctions
Coordinating conjunctions link two independent clauses. The phrases ر/phir bʰī/ even so and
@ /tō bʰī/ or /tab bʰī/ nevertheless in Section 7.2.1.3.2 are coordinating conjunctions, as is
/jæsē hī/ as soon as (see Section 7.2.1.2 above).
اور/ɔr/ is most commonly used to connect two words, phrases, or clauses which play the same role
in the sentence.
اور/ɔr/ may also be used as a qualifier, meaning more (see Section 7.2.1.3.2). In this usage, the
noun it modifies may be omitted.
132
Miscellaneous
(7.29) ! اور
ɔr kʰān-ā l-ō
and food-M.SG take-IMP.2.PL
‘Take more food!’
(7.30) ! اور
ɔr l-ō
and take-IMP.2.PL
‘Take more (food)!’
7.2.2.1.2 /yā/or
Note that ی: /karī/ is a dialectal form; it is rare in written Urdu, though common in the speech
of some dialects.
133
Miscellaneous
A clause beginning with ہ/balkeh/ means on the contrary when preceded by a negative clause.
It may also redirect or reinforce the meaning of an affirmative clause.
ہC /jab keh/ introduces a clause contemporaneous with the action of a preceding clause:
7.2.2.2.1 اگ/agar/ if
Note that /cācā/ is a common variant; the textbook form would be /cacā/.
134
Miscellaneous
اگ/agar/ if, introduces the condition clause of a conditional sentence; see Section 8.3.2.2.2 for an
account of conditionals.
7.2.2.2.2 ہC /tā keh/ so that, and ہCطa /bašartekeh/ provided that
ہC /tā keh/ so that, and ہCطa /bašartekeh/ provided that, introduce clauses in the subjunctive mood.
/jab tak/ followed by a negative introduces a relative subordinate clause; the construction
as a whole means until.
jab tak abbā xud yahāṁ nah ā-ē mæṁ tumh-ēṁ kucʰ
when until dad self here NEG come-PP.M.PL I you-ACC something
nahīṁ d-ūṁ g-ī
NEG give-SBJV.1.SG FUT-F.SG
‘Until Dad is around (arrives), I’m not giving you anything.’
ہC /keh/ may replace the causal conjunction ہ /kyūṁ keh/ (see Section 7.2.2.5) and the subor-
dinating conjunction ہC /tā keh/ (see Section 7.2.2.2.2), particularly in poetry.
135
Miscellaneous
ہC /keh/ may replace /yā/ or, in questions presenting an either/or choice.
ہC /keh/ may replace /jō/, in its sense of if, in linking two contrasting clauses; this is a Persian
usage, and relatively uncommon in Urdu:
136
Miscellaneous
ہC /keh/ introduces clauses which are the direct objects of verbs of speech or perception: e.g.,
/kahnā/ to say; بڑ/paṛʰnā/ to read; and /sunnā/ to hear.
Following a verb of speech, ہC /keh/ may introduce a direct or indirect quotation:
Following a verb of perception or apprehension, ہC /keh/ introduces a coordinate clause detailing
what has been perceived or thought:
ہC /keh/ may also introduce clauses elaborating on the type or kind of situation laid out in the
main clause:
137
Miscellaneous
ہC /keh/ also introduces subordinate clauses describing the results of action or circumstances in
the main clause. The main clauses of such sentences will often contain a modifier of extent, e.g.,
one of the 6 ا/itnā/ series.
Many phrases in the subjunctive mood begin in ہC /keh/; these typically express possibility, ne-
cessity, or hope.
138
Miscellaneous
Correlative conjunctions are a specific kind of coordinating conjunction; they link two clauses of
equal importance and equivalent function. They may be inclusive (as with both...and) or exclusive
(as with either...or).
7.2.2.4.1 Inclusive
The inclusive correlative conjunction, ... /bʰi...bʰi/ both...and, links two words, phrases, or
clauses. If it is linking phrases or clauses with the same verb, the verb may often be omitted from
the second part.
7.2.2.4.2 Exclusive
/yā/, or (see Section 7.2.1.2), may be used in both halves of a correlative clause; in this usage,
it means either...or.
بہ۔۔۔بہ/nah...nah/, used the same way, is the negative equivalent of this construction: neither...nor.
139
Miscellaneous
8... 8 /kyā...kyā/ whether...or, is a contrastive conjunction. It links two adjectives or nouns within
a subordinate clause, which describe or expand on the subject or object of the main clause.
ہ /kyūṁ keh/ because, introduces a reason clause: the answer to a direct question.
Reason clauses in ہ /kyūṁ keh/ come at the end of the sentence, following their result clauses.
Reason clauses may also be introduced by ہ/cūṁ keh/ since, which is usually paired with اس
/is liē/ in the result clause. Reason clauses in /cūṁ keh/ come at the beginning of the sentence .
140
Miscellaneous
Concessive clauses introduced by اگ ہ/agarceh/ although, tend to be followed by ر /pʰir bʰī/
nevertheless, introducing the conclusion.
Concessive clauses introduced by ﻻ ہ/hālāṁ keh/ even though, may precede or follow their main
clauses:
141
Miscellaneous
Sentential particles do not modify a sentence grammatically; rather, they express the speaker’s
intent or attitude. They may introduce or conclude a statement. Many are secondary usages of
other words, generally adverbs and question words.
ں/kyōṁ / (lit. why). As a particle, it solicits the hearer’s reaction to the sentence; it may be
translated as well. It is very informal, and used only when addressing a person.
ںبہ/kyōṁ nah/ (lit. why not), offers a suggestion; it may be translated as how about.
(7.65) ۓ ب ؟ ںبہ
ّ ا/accʰā/ means good; used as an introductory particle, it expresses surprise and sometimes
disapproval. It may be translated by a variety of English interjections.
7.2.3.2 Tags
/sahī/ very well, all right. This is a concessive tag; it grants the position taken in the preceding
statement.
142
Miscellaneous
(7.67) و
It may also optionally follow the particles @ /tō/, /nah/, and /hī/.
ُ
(7.68) @ رے @ا @ رے س8ہC @ @
(Note that this usage of the locative س /ke pās/ for intangible possession is unusual (see
Section 5.1.1.1.2 for more on the types of possession). This is probably due to English influence.)
/nā/, as a tag, turns a statement into a yes/no question, expecting an affirmative answer; it may
be translated as isn’t it?, right?, or simply eh?
When appended to a courteous request ending in the future form گ/gā/, /nā/ transforms the
sentence into a courteous question. Appended to other requests, usually at the /tum/ politeness
level, it acts as an emphatic.
tum ā-ō gē nā
you come-SBJV.2.PL FUT.M.PL TAG
‘You will come, wont you?’
143
Miscellaneous
As a sentential particle, the interrogative 8 /kyā/ signals a yes-or-no question. In this usage,
8/kyā/ can occur either at the the beginning of the sentence or as a tag, although it is most
commonly used at the beginning. When used as a tag at the end of a sentence, it may express
doubt or challenge.
(Among older speakers, and in more formal speech, this locative construction in 8__ س/kyā__pās/
would not be used of intangible objects such as knowledge, time, etc. Instead, the expected form
would be ب 8 /kyā tumhēṁ ilm hæ/.)
There are three negative particles in Urdu: بہ/nah/, ب/nahiṁ /, and /mat/. The first two can
be used in any context (although there are complicated and evolving rules on which is preferred
in any given situation), but /mat/ is restricted to commands and subjunctives. When used in
commands, the three follow a scale of politeness, with بہ/nah/ being perceived as neutral, ب
/nahiṁ / as more insistent, and /mat/ as even emphatic, possibly with a harsh connotation. بہ
/nah/ therefore is used with polite imperative forms; /mat/ with @ /tū/ and /tum/ forms.
ب/nahiṁ / in commands is a colloquial form, and more common among younger speakers. All
three negative particles can be used with the subjunctive when it is used as a command form.
144
Miscellaneous
(7.74) ! 9 اببہد5اس ت
(7.76) ! و: ز دہ صہ ب
The particle /sā/ forms adjectival phrases from adjectives, nouns, and pronouns; see Sec-
tion 6.6.1.
7.2.5.2 Interjections
The vocative case endings و/ō/ and ے/ē/ attach to nouns, while the vocative interjection ارے
/arē/ precedes names and noun phrases in direct address. و/ō/ and ے/ē/ may be translated
with O, and ارے/arē/ 2 with hey.
2
Although grammatical descriptions list a feminine version, اری/arī/; in practice, it is not used much, and the
masculine form is used for feminine vocatives as well.
145
Miscellaneous
! ڑ/laṛki-ō/ O girls!
! ارے ڑ/arē laṛk-ē/ Hey boy!
! ارے ڑ/arē laṛki-ō/ Hey girls!
These particles may introduce a sentence, or may stand alone. The more common free interjections
include واہ/vāh/ and ش/šābāš/, both meaning bravo!; او/ō hō/ oh no!; اؤی/ūī/ ouch! and ۓ
/hāē/ alas, both mostly restricted to women’s speech; ﷲ/bismillāh/ in God’s name; and ان ء
ﷲ/inšāallāh/ God willing.
Like many other South Asian languages, Urdu makes prolific use of repetition of words or parts
of words to express such notions as plurality, generality, repetition, and so on. In this grammar,
we refer to the simple repeating or doubling of whole words as repetition and to the repeating of
parts of words as reduplication.
(7.78) گآۓ ب ں ڑ ڑ
As is seen across the South Asian linguistic area, Urdu has a construction in which a word can be
echoed by a nonsense word to give the first word a general sense. The second word is formed by
changing the initial consonant, usually to a /v/ or a /š/; or, in the case of vowel-initial words, by
adding the /v/ or /š/ to the beginning of the word.3 Reduplication occurs most often in colloquial
speech; it is considered highly informal and is rarely encountered in written Urdu. It is extremely
productive and can be used with loanwords as well as native Urdu words. It is not used with
personal names.
3
Note that use of /š/ in echo words comes from Panjabi, and such forms are regarded as Panjabi, but they are not
uncommon in Urdu due to frequent Urdu-Panjabi code-switching and other contact phenomena.
146
Miscellaneous
ۓواۓ/cāē vāē/ tea and the stuff that goes with it (< ۓ/cāē/ tea)
ۓ ۓ/cāē šāē/ tea and the stuff that goes with it
• @ ن @ ن/kɔn kɔn/ which (various) people? (See “Notes” in Section 4.6 for more discussion of this
construction.)
(7.80) ک @ ن@ نآ
kal kɔn kɔn ā-yā tʰā
yesterday who who come-PP.M.SG be.PST.M.SG
‘Which (various) people came yesterday?’
147
Miscellaneous
(7.84) ھبہ ھ@ گ
kʰā-n-ē kē.liyē kucʰ nah kucʰ tō hō gā
eat-INF-OBL for some NEG some TOP be FUT.M.SG
‘There must be something (or other) to eat.’
(7.85) ھبہ ھ@ گ ا ری ب
almārī mēṁ pehn-n-ē kē.liyē kucʰ nah kucʰ tō hō gā
closet in wear-INF-OBL for some NEG some TOP be FUT.M.SG
‘There must be something or other to wear in the closet.’
Like nouns and pronouns (see Section 7.3.1 and Section 7.3.2), adjectives may be repeated or
reduplicated (repeated in part or with alterations).
Repetition of an adjective, along with any inflectional endings that would normally appear on it,
gives it a meaning of plurality or multiplicity:
ّ9 و و/cʰot ̣ē cʰot ̣ē baccē/ many small children; small children all over the place (lit. small
small children)
Adjective repetition may also carry affective meanings--specifically, a sense of goodness or appro-
priateness.
ا ا /t ̣hand̩ā t ̩hand̩ā pānī/ nice cold water (lit. cold cold water)
The phrase assumes that cold is a positive attribute of water-- that we are speaking of drinking
and not, say, of bathing-- and the reduplication intensifies the positive association. These shades
of meaning are not entirely predictable, and will often need to be learned phrase by phrase.
Adjective reduplication involves an alteration of the vowel, as in ک / t ̩hīk t ̩hāk/ all right,
from / t ̩hīk/ right. This echo element generalizes the meaning of the reduplicated word.
148
Miscellaneous
Participles can be repeated in Urdu to show continuing or repeated action, with nuances of mean-
ing expressed by the particular sort of participle that is being repeated.
Repeating an imperfective participle conveys two things about the action of the participle: first,
that it is continuous or repeated, and second, that its inception precedes the action of the main
verb. Often the action of the main verb either is the result of the previous action or else somehow
caused that action to end or prevented it from being completely accomplished. Generally a re-
peated imperfect participle does not have to agree with its subject; instead, the masculine oblique
149
Miscellaneous
singular form of the participle is used, regardless of the subject’s gender.4 However, if the partici-
ple and main verb share the same subject, and the subject is in the direct case, the participle may
optionally agree with its subject.
Two semantically related imperfective participles may be used one after the other in a similar
way:
4
This especially true of Delhi Urdu.
150
Miscellaneous
Perfective participles can also be repeated to convey an action that is repeated, but not continued
(since perfective participles signify completed action). As with repeated imperfective participles,
the most common form, regardless of the subject, is masculine oblique singular, unless the main
verb is in a perfective tense and the subject in the direct case. In these instances, the participle
generally agrees with the subject.
151
Verbs
Chapter 8
Verbs
8.1.1 Overview
The various inflected verb forms and verb constructions of Urdu are created by adding certain
elements to one of the following four basic forms:
• the stem
• the infinitive
These basic forms combine with various suffixes and auxiliary verbs1 to produce a wide range of
tense and aspect forms of the Urdu verbal system.2
In Urdu, tense is indicated through auxiliary verbs (or, in the case of the future, through a suffix).
Among the tenses, each of the three aspects is associated with one basic verbal form, upon which
a full verbal construction, with tense and aspect, is based:
• stem (along with obligatory perfective participle of ر/rahnā/ to remain) → durative aspect (the
progressive tenses)
152
Verbs
Verbs are inflected to agree in number and person with their associated nouns (usually the subject,
but in the case of perfect tenses of transitive verbs, the object). As discussed in Section 4.2.1,
subjects in the third person that have a singular referent can take a plural verb, to indicate a
higher degree of respect.
(8.1) ر:م5
versus
<!--Verb features
In addition to the features below, verbs may be inflected for the
gender and number of their subjects (or sometimes objects); those
features are defined for nouns. Also, infinitives can be inflected for
case, since they are in essence nouns.
-->
<Fs:ClosedFeatureDefn name='Causative' id='fdefnCausative'>
<!--This is an ad hoc feature, but it's unclear what the
morphosyntactic structure of a causative should be.-->
<Fs:SymbolicFeatureValueDefn symbol='single' id='fvSingle'/>
<Fs:SymbolicFeatureValueDefn symbol='double' id='fvDouble'/>
</Fs:ClosedFeatureDefn>
153
Verbs
As described below, verbs can also be inflected to agree in gender and number with nouns; the
formal grammar uses the noun features described in Chapter 3 for agreement on verbs.
The stem is the base, or bare form, of the verb to which suffixes are added. Generally speaking,
the stem of a given verb can be inferred from the infinitive: taking away the infinitival suffix
/-nā/ leaves the stem. Urdu stems have three basic forms.
Formation:
• plain stem:3
plain stem + ∅ (i.e., no suffix)
• causative stem:4
plain stem + ا/-ā/ or ﻻ/-lā/
154
Verbs
: ھ
Plain stem /kar/ /sun/ /bæt ̣ʰ/
to do [something] to hear to sit
ا:
/karā/
Causative stem to cause [something] /sunā/
to tell (lit. to cause /bit ̣ʰā/
to be done, to have to make [someone] sit
someone to hear)
[something] done
وا: وا
Double causative /karvā/ /sunvā/ /bit ̣ʰvā/
stem to cause [someone] to to cause [someone] to to get someone to
do [something] tell make someone sit
Formation:
stem + ~ ~ ب ~ /-tā~ -tē ~ -tī ~ -tīṁ /
The suffix has variants because imperfective participles are adjectival in form and are therefore
inflected to agree in gender and number with their noun or pronoun complements. Usually, just
as with adjectives (see Table 6.1 for adjectival endings), there is no distinction between the sin-
gular and plural in the feminine: the feminine suffix /-tī/ is used for both singular and plural
imperfective participles. But where an auxiliary is not present (which happens with irrealis con-
ditionals with the imperfective participle and negative sentences of the present imperfect), the
plural marking that would have appeared on an auxiliary is shifted to the feminine participle, and
therefore the suffix /-tīṁ / is used instead for feminine plural imperfective participles.
Table 8.2 shows suffixes of the imperfective participle.
Masculine
/-tā/ /-tē/
ب
Feminine
/-tī/ /-tīṁ /
155
Verbs
The imperfective is used for actions which are incomplete, either because they have not yet been
finished, or because they are habitual or recurring.
Examples:
Formation:
stem + ا~ے ~ی~ب/-ā ~ -ē ~ -ī ~ -īṁ /
Like the imperfective participle, it is inflected adjectivally, to agree in gender and number with
its noun or pronoun complements.
Vowel-final stems combine with the ا/-ā/ and ی/-ī/ suffixes in the following ways:
• a + ā → ayā :
• i+ī→īی:
/jānā/ to go, perfective stem گ/ga-/; perfective participle گ/gayā/ ( ب،ی، ے/-ē, -ī, -īṁ /)
: /karnā/ to do, perfective stem ک/ki-/; perfective participle 8 /kiyā/ ( ب،ی، ے/-ē, -ī, -īṁ /)5
5
There is a dialectal form of the perfective participle, ا: /karā/, that is heard instead of 8 /kiyā/.
156
Verbs
6 د/dēnā/ to give, perfective stem د/di-/; perfective participle د/diyā/ ( ب،ی، ے/-ē, -ī, -īṁ /)
/lēnā/ to take, perfective stem ل/li-/; perfective participle /liyā/ ( ب،ی، ے/-ē, -ī, -īṁ /)
The perfective participle is used for the perfective aspect--completed or finished actions, and ac-
tions that occur only once. See Section 8.5 below for uses of the perfective.
Formation:
stem + /-nā/6
The infinitive is the citation form of the verb, which means that it is the form used when speaking
about that verb, as for example when answering the question, “How do you say eat in Urdu?”
Formally it is a verbal noun, occurring in the direct case as the subject of a verb or in the oblique
case with postpositions. When used this way, it is inflected as a singular masculine marked noun:
/-nā/ for the direct case and /-nē/ or the oblique. See Section 8.6 for uses of the infinitive.
Table 8.3 shows examples of infinitives.
:
From plain stem /karnā/ /sunnā/ /bæt ̣ʰnā/
to do [something] to hear to sit
ا:
From a causative /karānā/
to cause [something] /sunānā/ /bit ̣ʰānā/
stem to tell (lit. to cause to make [someone] to
to be done, to have
someone to hear) sit
[something] done
The verb /hōnā/ to be, to become is used as both a main verb and an auxiliary verb. As an
auxiliary verb, it is the basis of all the complex verbal constructions of Urdu. Its tense determines
the tense of verbal constructions in which it appears.
Since /hōnā/ is irregular and is used in so many verbal constructions, the relevant paradigms
are given here; the formal grammar fragment for this verb is given at the end of this section.
6
“Stem” here means any of the three types of stem.
157
Verbs
The present tense of /hōnā/ is used in present tense constructions. These forms agree with
their subject in person and number. Negatives are formed with the word ب/nahīṁ /, as they
are in all the indicative forms.
Table 8.4 shows the conjugation of the present tense form of the verb /hōnā/ to be.
2nd person
@ /tū hæ/ you are
singular
The simple past tense of /hōnā/ is used in past tense constructions. These forms are participial
in origin; as such, they agree with their subject in number and gender, but not person. The variants
within each entry represent masculine and feminine forms respectively.
The negative of the past tense of /hōnā/ is generally formed with ب/nahīṁ /.
Table 8.5 shows the conjugation of the past tense form of the verb /hōnā/ to be.
The subjunctive mood of /hōnā/ is used to refer to unrealized actions, for example in conditional
constructions, . These forms agree with their subject in person and number.
Table 8.6 shows the conjugation of the subjunctive mood form of the verb /hōnā/ to be.
In the subjunctive, the negative is formed with the particle بہ/nah/.
158
Verbs
Formation:
subjunctive + ~ ~ گ/gā ~ gē ~ gī/
In addition to indicating future time, the future tense of /hōnā/ is used in presumptive construc-
tions, to describe actions predicted to occur or assumed to have occurred. It is formed by adding
، ، 5 /gā ~ gē ~ gī/ to the subjunctive. These suffixes are usually written separately, but can
also be written attached to the preceding subjunctive form, and the alternate joined forms occur
often enough that we have included them in the table below.7 As with the past tense, these forms
agree with their subject in gender and number, and the variants in the table represent masculine
and feminine forms respectively. The negative is formed with /nahīṁ /.
Table 8.7 shows the conjugation of the future tense form of the verb /hōnā/ to be.
7
Note that when the endings are attached to the verb in writing, the nūn γunnah becomes a medial consonant and
consequently is represented as a full consonantal nūn, although this difference in spelling does not reflect a difference
in pronunciation: regardless of spelling, one still hears only a nasalized vowel + /g/, not a vowel + full consonantal
nasal + /g/.
159
Verbs
2nd person
@ /tū hō/ you may be/should be
singular
160
Verbs
161
Verbs
<Fs:SymbolicFeatureValue defn='fvIndicative'/>
</Fs:FeatureStructure>
</Irr:WordForm>
<Irr:WordForm inflectionalGloss='-3SgPres'>
<Ph:Form spelling=' '/>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fv3'/>
<Fs:SymbolicFeatureValue defn='fvSingular'/>
<Fs:SymbolicFeatureValue defn='fvPresent'/>
<Fs:SymbolicFeatureValue defn='fvIndicative'/>
</Fs:FeatureStructure>
</Irr:WordForm>
<Irr:WordForm inflectionalGloss='-1PlPres'>
<Ph:Form spelling=' ' ب/>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fv1'/>
<Fs:SymbolicFeatureValue defn='fvPlural'/>
<Fs:SymbolicFeatureValue defn='fvPresent'/>
<Fs:SymbolicFeatureValue defn='fvIndicative'/>
</Fs:FeatureStructure>
</Irr:WordForm>
<Irr:WordForm inflectionalGloss='-2PlInfPres'>
<Ph:Form spelling=' '/>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fv2'/>
<Fs:SymbolicFeatureValue defn='fvPlural'/>
<Fs:SymbolicFeatureValue defn='fvInformal'/>
<Fs:SymbolicFeatureValue defn='fvPresent'/>
<Fs:SymbolicFeatureValue defn='fvIndicative'/>
</Fs:FeatureStructure>
</Irr:WordForm>
<Irr:WordForm inflectionalGloss='-2PlFormPres'>
<Ph:Form spelling=' ' ب/>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fv2'/>
<Fs:SymbolicFeatureValue defn='fvPlural'/>
<Fs:SymbolicFeatureValue defn='fvFormal'/>
<Fs:SymbolicFeatureValue defn='fvPresent'/>
<Fs:SymbolicFeatureValue defn='fvIndicative'/>
</Fs:FeatureStructure>
</Irr:WordForm>
<Irr:WordForm inflectionalGloss='-3PlPres'>
<Ph:Form spelling=' ' ب/>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fv3'/>
<Fs:SymbolicFeatureValue defn='fvPlural'/>
<Fs:SymbolicFeatureValue defn='fvPresent'/>
162
Verbs
<Fs:SymbolicFeatureValue defn='fvIndicative'/>
</Fs:FeatureStructure>
</Irr:WordForm>
163
Verbs
<Fs:SymbolicFeatureValue defn='fvSingular'/>
<Fs:SymbolicFeatureValue defn='fvSubjunctive'/>
</Fs:FeatureStructure>
</Irr:WordForm>
<Irr:WordForm inflectionalGloss='-2SgSubjunc'>
<Ph:Form spelling=' '/>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fv2'/>
<Fs:SymbolicFeatureValue defn='fvSingular'/>
<Fs:SymbolicFeatureValue defn='fvSubjunctive'/>
</Fs:FeatureStructure>
</Irr:WordForm>
<Irr:WordForm inflectionalGloss='-3SgSubjunc'>
<Ph:Form spelling=' '/>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fv3'/>
<Fs:SymbolicFeatureValue defn='fvSingular'/>
<Fs:SymbolicFeatureValue defn='fvSubjunctive'/>
</Fs:FeatureStructure>
</Irr:WordForm>
<Irr:WordForm inflectionalGloss='-1PlSubjunc'>
<Ph:Form spelling='' ں/>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fv1'/>
<Fs:SymbolicFeatureValue defn='fvPlural'/>
<Fs:SymbolicFeatureValue defn='fvSubjunctive'/>
</Fs:FeatureStructure>
</Irr:WordForm>
<Irr:WordForm inflectionalGloss='-2PlInfSubjunc'>
<Ph:Form spelling=' '/>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fv2'/>
<Fs:SymbolicFeatureValue defn='fvPlural'/>
<Fs:SymbolicFeatureValue defn='fvInformal'/>
<Fs:SymbolicFeatureValue defn='fvSubjunctive'/>
</Fs:FeatureStructure>
</Irr:WordForm>
<Irr:WordForm inflectionalGloss='-2PlFormSubjunc'>
<Ph:Form spelling='' ں/>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fv2'/>
<Fs:SymbolicFeatureValue defn='fvPlural'/>
<Fs:SymbolicFeatureValue defn='fvFormal'/>
<Fs:SymbolicFeatureValue defn='fvSubjunctive'/>
</Fs:FeatureStructure>
</Irr:WordForm>
<Irr:WordForm inflectionalGloss='-3PlSubjunc'>
164
Verbs
8.3.1 Imperatives
There are three levels of imperative, or request, forms in Urdu, not counting the impersonal request
with the infinitive (see Section 8.6.2). These three correspond to the three levels of familiarity
among the pronouns @ /tū/, /tum/, and آپ/āp/.
See Section 7.2.4 for how to negate imperatives; i.e., how to express prohibitions.
8.3.1.1 Intimate
Formation:
stem + ∅
The least formal of the three levels of request form corresponds to the intimate second-person
pronoun @ /tū/. It consists of the bare verb stem alone. Imperatives corresponding to @ /tū/ are
used only when addressing very small children, pets, and God. Note that when that when used in
contexts inappropriate for intimate forms, these imperatives are considered extremely rude and
demeaning and, in those instances, are meant to belittle or chastise.
! ( @) /(tū) sun/ Listen!
8.3.1.2 Familiar
Formation:
stem + و/-ō/8
This mid-level imperative corresponds to the pronoun /tum/. In keeping with the use of
/tum/ for both singular and plural subjects, this form can be used in singular and plural contexts.
! ) ( ب/(tum) sunō/ Listen!
8
Equivalent to the second person plural subjunctive; see Section 8.3.2.1 below.
165
Verbs
! ) ( ؤ/(tum) sunāō/ Cause (me) to listen! (e.g., play some music for me so that I can hear)
8.3.1.3 Formal/Polite
Formation:9
stem + /-iyē/
This ending is spelled ۓ/-iē/ ~ /-īē/ ~ /iyē/ when added to stems ending in consonants.
:
/karnā/ →
/kījie/
to do
6د 9د
/dēnā/ →
/dījie/
to give
/lēnā/ →
/lījie/
to eat
/pīnā/ →
/pījie/
to drink
9
The subjunctive can also be used to express a formal imperative with a slightly less formal connotation than ۓ
/-iye/, as in let’s go; why don’t you have a seat, etc. (See below, Section 8.3.2.2.1.)
166
Verbs
8.3.2 Subjunctive
Formation:
stem + personal endings (see Table 8.8 below)
Both regular and irregular subjunctive verbs agree with their subjects in person and number.
2nd person
ے: @ /tū karē/ you may do
singular
The verb /hōnā/ to be is irregular in the subjunctive; see Section 8.2.3 above. The verbs 6د
/dēnā/ to give and /lēnā/ to take also follow an irregular conjugation:
Table 8.9 shows irregular subjunctive conjugation in the verb 6 د/dēnā/ to give; Table 8.10 shows
irregular subjunctive conjugation in the verb /lēnā/ to take.
The subjunctive mood is used for unrealized actions or events. It is required by certain phrases
and adverbs, particularly those expressing doubt or contingency, and by some types of dependent
167
Verbs
2nd person
@ دے /tū dē/ you may give
singular
2nd person
@ /tū lē/ you may take
singular
168
Verbs
clauses.10 Subjunctive verbs are often translated with may, might, or should. They aways take بہ
/nah/ in the negative.
Some words or phrases that require or frequently take the subjunctive are listed in the following
table. Examples of each in a sentence can be found in the next two sections.
/šāyad/ maybe
introduces subordinate
Verb + ہC Verb + /keh/
clause
• ask permission:
10
A dependent, or subordinate, clause is a part of a sentence that has a subject and a verb, but is dependent on
(subordinate to) the sentence’s main clause; that is, they could not stand alone. An example in English would be the
sentence if you practice every day, you will learn to play piano well, where if you practice every day is the subordinate
clause and you will learn to play piano well is the main clause. The subject of the main clause and the subject of the
dependent clause may be the same, as in the example above, or they may be different, as in when he went outside, the
sky was cloudy, where when he went outside is the dependent clause and the sky was cloudy the main clause.
11
Subjunctive not always required; in irrealis contexts, the simple imperfective participle is used.
169
Verbs
(8.7) ! ب:آپ
āp kar-ēṁ
you do-SBJV.2.PL
‘You should do it!’
170
Verbs
While not all subordinate clauses in Urdu require the subjunctive, many do. The subjunctive is
used when a subordinate clause expresses a contingent situation introduced by a subordinating
conjunction,13 including /keh/ that:
• اگ/agar/ if 14
• /jab tak/ + NEGATIVE unless, until (introduces a relative clause which may contain a
subjunctive or an indicative)
171
Verbs
172
Verbs
173
Verbs
Conditional sentences are if...then statements; in Urdu, they generally consist of a conditional
clause, introduced by اگ/agar/ if and a result clause, introduced by @ /tō/ then. اگ/agar/ is often
deleted, and may be replaced by /jab/ or /jō/. @ /tō/ is rarely deleted or replaced.
The mood and tense of the verb in the if clause determine the status of the condition--whether it
is fulfilled, and, if not, whether it may yet be. The tense of the verb in the result clause determines
the tense of the sentence as a whole.
The subjunctive can be used in the ‘if’ clauses of fulfillable conditions, to express actions whose
outcome is unknown--either because the action has not yet taken place, or because its occurrence
or non-occurrence is not known to the speaker:
(8.26) @ ن ر5اگ د
8.3.3 Future
Formation:
174
Verbs
As is clear from the above, the Urdu future tense is formed by adding the particle ~ ~ گ/gā ~
gē ~ gī/ to the subjunctive. Although they function as suffixes, these endings are usually written
separately from the first part of the verb. However, it is not uncommon to see them written
attached to the preceding subjunctive form. These alternate joined forms occur often enough
that we have included them in the table below.16 The future endings are participial and agree
with the subject in gender and number; the subjunctive inflection, already present on the verb,
shows agreement in person and number. Forms shown in the table are masculine and feminine
respectively.
Negatives in the future tense are formed with ب/nahīṁ /.
Table 8.12 shows the conjugation of the future tense form of the verb : /karnā/ to do.
The future in Urdu works much like English future tense verbs with will--it expresses actions yet
to take place.
Like the subjunctive, the future is used in condition clauses to describe actions of unknown out-
come:
175
Verbs
/mæṁ karūṁ gā ~
و:~ وںگ: ب karūṁ gā/
1st person singular I will do
و:~ وں: ب /mæṁ karūṁ gī ~
karūṁ gī/
/tū karē gā ~
2nd person :~ ےگ: @
karēgā/ you will do
singular :~ ے: @ /tū karē gī ~ karēgī/
/voh karē gā ~
:~ ےگ:وہ karēgā/ he/it will do
3rd person singular
:~ ے:وہ /voh karē gī ~ she/it will do
karēgī/
/tum karō gē ~
و:~ و:
2nd person plural karōge/
you will do
(informal) و:~ و: /tum karō gī ~
karōgī/
/āp karēṁ gē ~
6 :~
2nd person plural ب:آپ karēṁ ge/
you will do
(formal) 6 :~ ب:آپ /āp karēṁ gī ~
karēṁ gī/
/voh karēṁ gē ~
6 :~ ب:وہ karēṁ gē/
3rd person plural they will do
6 :~ ب:وہ /voh karēṁ gī ~
karēṁ gī/
176
Verbs
Formation:17
stem + : /kar/ [unconjugated] (for all verbs but : /karnā/)
(8.30) : ہ ل6 5
versus:
Formation:
stem + ~ر ب ~ر ر ~ر/rahā ~ rahī ~ rahē ~rahīṁ / + inflected auxiliary verb ( /hōnā/)
The progressive tenses describe actions or states which are in progress. (That is to say, they have
durative aspect.)
The word ر/rahā/ inflects like an adjective and agrees with the subject in gender and num-
ber.19 The auxiliary /hōnā/ agrees with the subject in number and person in all forms except
the progressive past and progressive irrealis, where, like ر/rahā/, it agrees in gender and number.
Its tense determines the tense of the entire construction.
17
In older writings, the stem by itself may occur as a conjunctive; however, this usage is considered incorrect in
modern standard Urdu.
18
Because in perfective tenses, the subject of a transitive verb takes the ergative postposition /nē/, and the verb
agrees with the direct object. (See Section 5.1.3 and below, Section 8.5, for more on ergative marking.)
19
Originally the perfective participle of the verb ر/rahnā/ to remain. The verb ر/rahnā/ is not usually used as the
main verb of a sentence in the progressive tenses: instead, speakers will use the present imperfect. When ر/rahnā/
177
Verbs
The progressive present is formed with the auxiliary /hōnā/ in the present tense. It describes
actions or states which are in progress at the time of speaking and may also be used for future
actions which can be considered to have already begun, or be imminent, as in English, I am going
to India tomorrow. Forms in the table are masculine and feminine respectively.
Negatives, although rare in this tense, are formed with ب /nahīṁ / and if they are used,
/hōnā/ is optional.
Table 8.13 shows the conjugation of the progressive present tense form of verb : /karnā/ to do.
ب ر:
/ham kar rahē hæṁ /
1st person plural we are doing
ب ر: /ham kar rahī hæṁ /
ر:
2nd person plural /tum kar rahē hō/
you are doing
(informal) ر: /tum kar rahī hō/
does occur in a progressive tense, it is to stress the temporary or limited duration of the action; for example, اند@ ں ب
سرہر ں دوس ب ا/in dinōṁ mēṁ mæṁ ek dōst kē pās rah rahā hūṁ / These days I am staying (temporarily) with
a friend.
178
Verbs
The progressive past is formed with the auxiliary /hōnā/ in the past tense. It describes actions
or states which were in progress at a point in the past (and which may or may not still be in
progress now.) Note that /hōnā/ as well as ر/rahā/ agrees in gender and number with the
subject; as before, forms in the table are masculine and feminine respectively.
Table 8.14 shows the conjugation of the progressive past tense form of verb : /karnā/ to do.
ر:
/ham kar rahē tʰē/
1st person plural we were doing
ب ر: /ham kar rahī tʰīṁ /
ر:
2nd person plural /tum kar rahē tʰē/
you were doing
(informal) ب ر: /tum kar rahī tʰīṁ /
179
Verbs
(Note also that the above sentence is an example of the third person plural being used to refer to
a singular person as a sign of respect.)
The progressive subjunctive is formed with the auxiliary /hōnā/ in the subjunctive or indicative,
and is introduced by one of the verbs, indirect constructions, or subordinating conjunctions in
Section 8.3.2.2.2. It describes progressive actions or states whose status at the time of speaking is
unknown or uncertain. As always with subjunctive verbs, the negative is formed with بہ/na/. As
before, forms in the table are masculine and feminine respectively.
Table 8.15 shows the conjugation of the progressive subjunctive form of verb : /karnā/ to do.
180
Verbs
ں ر:
/ham kar rahē hōṁ / we may be doing, if
1st person plural
ں ر: /ham kar rahī hōṁ / we are doing
ر:
2nd person plural /tum kar rahē hō/ you may be doing, if
(informal) ر: /tum kar rahī hō/ you are doing
2nd person plural ں ر:آپ /āp kar rahē hōṁ / you may be doing, if
(formal) ں ر:آپ /āp kar rahī hōṁ / you are doing
181
Verbs
The progressive presumptive is formed with the auxiliary /hōnā/ in the future. It describes
actions or states which the speaker presumes to be occurring. Forms in the table are masculine
and feminine respectively.
Table 8.16 shows the conjugation of the progressive presumptive form of the verb : /karnā/ to
do.
ں ر:
/ham kar rahē hōṁ gē/
1st person plural we must be doing
ں ر: /ham kar rahī hōṁ gī/
ر:
2nd person plural /tum kar rahē hō gē/
you must be doing
(informal) ر: /tum kar rahī hō gī/
2nd person plural ں ر:آپ /āp kar rahē hōṁ gē/
you must be doing
(formal) ں ر:آپ /āp kar rahī hōṁ gī/
voh sō rah-ā hō gā
he sleep PROG-M.SG be.SBJV.3.SG FUT.M.SG
‘He must be sleeping. (in another context, “He will be sleeping”)’
182
Verbs
The progressive irrealis is formed with the imperfective participle of the auxiliary /hōnā/ (See
Section 8.1.3). It is used to describe contrary-to-fact actions or states, which would be in progress
at the time of speaking, if only they were true.
Table 8.17 shows the conjugation of the progressive irrealis form of the verb : /karnā/ to do.
ر:
/ham kar rahē hōtē/
1st person plural (if) we were doing
ب ر: /ham kar rahī hōtīṁ /
ر:
2nd person plural /tum kar rahē hōtē/
(if) you were doing
(informal) ب ر: /tum kar rahī hōtīṁ /
183
Verbs
As described in Section 8.1.3 above, the imperfective participle consists of the verb stem plus
the suffix ~ ~ ب ~ /-tā ~ -tē ~ tī ~ tīṁ /, which inflects like an ordinary marked adjective,
agreeing with the subject in gender and number.
Verb structures based on the imperfective participle describe actions or states which occur gen-
erally or regularly, are ongoing, or are in some way incomplete.20 This is known as imperfective
aspect. These verb constructions are formed from the imperfective participle followed by an in-
flected form of the auxiliary verb /hōnā/. The tense of the auxiliary verb determines the tense
of the construction as a whole.
8.4.1.1 Irrealis
The imperfective participle may be used on its own, with no auxiliary, in the two clauses of a
conditional sentence, to describe an unfulfilled or unfulfillable condition.
(8.42) @ ب :ن
The imperfective participle may also occur by itself to express impossible conditions.
20
Including when the bare participle is used for an irrealis situation, which is incomplete in the sense that it did not
or will not happen; see Section 8.4.1.1 just below.
184
Verbs
Narrative passages using the past imperfect will typically use the full verb + auxiliary construction
for the first occurrence, then omit the auxiliary for the following verbs, using the imperfective
participle by itself. Grammarians call such dropping of an expected form ellipsis.
(8.45) شبہ5ہC اورسو روزب وں@ دa ۔ ر a ۔و ںا ب ےر غب ب ا ز
ے ں ب آ ب ۔a ب ے
Formation:
imperfective participle + /hōnā/ (present tense)
The present imperfect describes actions or states which occur generally or regularly, at or around
the time of speaking. It can also be used to describe events in the near or imminent future as in
English, where we might say He flies to Minneapolis tomorrow.
The negative uses ب /nahīṁ / and may omit the auxiliary verb. If the auxiliary is dropped, the
feminine plural suffix ب/-tīṁ / is substituted for /-tī/.
Table 8.18 shows the conjugation of the imperfect present tense form of the verb : /karnā/ to do.
185
Verbs
ب :
/ham kartē hæṁ /
1st person plural we do
ب : /ham kartī hæṁ /
:
2nd person plural /tum kartē hō/
you do
(informal) : /tum kartī hō/
Formation:
imperfective participle + /hōnā/ (past tense)
The past imperfect describes actions or states which occurred generally or regularly in the past.
Depending on context, it may be translated in English with used to, X-ed, or was/were X-ing.
186
Verbs
The negative uses ب /nahīṁ /. Note that the auxiliary agrees with the subject in gender and
number.
Table 8.19 shows the conjugation of the imperfect past tense form of the verb : /karnā/ to do.
:
/ham kartē tʰē/
1st person plural we used to do
ب : /ham kartī tʰīṁ /
:
2nd person plural /tum kartē tʰē/
you used to do
(informal) ب : /tum kartī tʰīṁ /
Formation:
187
Verbs
ں :
/ham kartē hōṁ /
1st person plural we may do
ں : /ham kartī hōṁ /
:
2nd person plural /tum kartē hō/
you may do
(informal) : /tum kartī hō/
188
Verbs
Formation:
imperfective participle + /hōnā/ (future)
The presumptive imperfect describes actions or states which are presumed to occur regularly in
the present.
Table 8.21 shows the conjugation of the presumptive imperfect form of the verb : /karnā/ to do.
Formation:
imperfective participle + /hōnā/ (imperfective participle)
The irrealis imperfect describes contrary-to-fact habitual actions or states-- actions which would
occur regularly, but are unrealized or impossible.
Table 8.22 shows the conjugation of the irrealis imperfect form of the verb : /karnā/ to do.
189
Verbs
I must do/must be
ںگ : ب /mæṁ kartā hūṁ gā/
1st person singular doing/will do/will be
ں : ب /mæṁ kartī hūṁ gī/
doing
190
Verbs
:
/ham kartē hotē/
1st person plural (if) we did
ب : /ham kartī hōtīṁ /
:
2nd person plural /tum kartē hotē/
(if) you did
(informal) ب : /tum kartī hōtīṁ /
191
Verbs
8.4.2.6 As Adjectives
Imperfective participles may also be inflected and used as adjectives, as in English flowing river,
rising damp. Frequently it is followed by the perfective participle of /hōnā/ in this usage. It
agrees with the noun it modifies.
(8.56) @ رترو 5
As described in Section 8.1.4 above, the perfective participle consists of the stem plus the suffix ا
/-ā/, which inflects like an adjective.
Formation:
perfective participle + ∅
The simple perfective is formed with the perfective participle alone, inflected appropriately for
agreement. It can refer to events in the past or the immediate future, or, very often, to unrealized
events (i.e., realis conditionals).
• Past
(8.57) فد ب
mæṁ nē ɣilāf dʰō-yā
I ERG cover.DIR wash-PP.M.SG
‘I washed the (pillow) cover.’
192
Verbs
• Immediate future
(8.58) ب ا آ
mæṁ abʰī ā-yā
I.DIR now come-PP.M.SG
‘I am coming right now/right away.’
• Realis conditional
(8.59) گ ھ آ @ وہآپ اگ
agar salīm ā-yā tō voh āp k-ē sāth cal-ē
if Saleem come-PP.M.SG EMPH he you POSS.OBL side walk-SBJV.3.SG
gā
FUT.M.SG
‘If Saleem comes, he’ll go with you.’
In the condition clause of a realis conditional sentence, the perfective participle indicates that the
speaker perceives a high degree of probability that the event will occur.
Formation:
perfective participle + inflected form of /jānā/
In a construction with /jānā/, the perfective participle expresses either the passive or incapacity.
8.5.1.2.1 Passive
With transitive verbs, this construction can give a passive sense. With the following exception,
both participle and /jānā/ agree with the subject, which, as in English, is the direct object in
the corresponding active, transitive sentence. However, if the “subject” of the passive sentence is
marked with @ /kō/, both parts of the verb revert to the default masculine singular marking.
If an agent or instrument of the action is expressed, it is followed by the postpositional phrases
ذر> ہ /kē zariē/, ھ /kē hath/ (preferred for human agents), or /sē/.
193
Verbs
Note that the correspondence between active and passive counterparts is not automatic. In any
given case, one or the other is more appropriate, and they are not interchangeable.
8.5.1.2.2 Incapacity
The same construction can be used in the negative to express incapacity. This usage occurs with
both transitive and intransitive verbs. The logical subject of the (non-)action is marked with
/sē/ (there is no grammatical subject if the verb is intransitive), and the verb inflection agrees
with the grammatical subject. If the verb involved is intransitive, there is no grammatical subject,
and the verb reverts to the default masculine singular.
194
Verbs
Together with an inflected form of the verb : /karnā/ to do, the perfective participle expresses
an action that is done iteratively; that is, on more than one occasion. The participle is in the
masculine singular,direct case.
The perfect tenses are formed from the perfective participle, followed by an inflected form of
the auxiliary /hōnā/, whose form determines the tense of the construction as a whole. The
perfective participle is inflected to agree with either the subject or the object, depending on the
verb. Intransitive verbs agree with the subject, transitive verbs with the direct object, unless it
195
Verbs
is marked with @ /kō/; while the subject of a transitive verb takes the “ergative” postposition
/nē/ and is therefore in the oblique case.
Perfect verb constructions describe events which occurred at one point of time in the past: that
is, they have perfective aspect.
Formation:
perfective participle + /hōnā/ (present)
The present perfect describes a state or action which is completed, but whose effects persist into
the present-- an action which created a lasting or lingering state. It therefore frequently refers to
a recently completed event. It can also be somewhat misleadingly called the immediate past, but
its functions are broader than that name implies.
Negatives are formed with ب/nahīṁ /.
Table 8.23 shows the conjugation of the perfect present tense form of /jānā/ to go.
ب
/ham gaē hæṁ /
1st person plural we have gone
ب /ham gaī hæṁ /
196
Verbs
(8.69) نگ5آ د
Formation:
perfective participle + /hōnā/ (past)
The past perfect, also called the remote past by some grammars, describes a state or action which
was completed in the past and whose effects no longer persist-- an action which created a state
that has now ended. It may also be used of actions completed in the remote past, prior to another
past action, or within a specific past timeframe.
Negatives are formed with ب/nahīṁ /.
Table 8.24 shows the conjugation of the perfect past tense form of the verb /jānā/ to go.
Formation:
perfective participle + /hōnā/ (subjunctive)
The perfect subjunctive, or past subjunctive, is introduced by one of the verbs, indirect construc-
tions, or subordinating conjunctions in Section 8.3.2.2.2. It describes actions or states which may
have occurred--and been completed--in the past.
As with all subjunctive tenses, the negative is formed with بہ/na/.
Table 8.25 shows the conjugation of the perfect subjunctive form of the verb /jānā/ to go.
197
Verbs
(8.72) ڑآ د ب
Formation:
perfective participle + /hōnā/ (future)
The presumptive perfect describes events which are presumed to have occurred--and been completed-
-in the past.
Table 8.26 shows the conjugation of the presumptive perfect form of the verb /jānā/ to go.
198
Verbs
ں
/ham gaē hōṁ /
1st person plural we may have gone
ں /ham gaī hōṁ /
(8.73) وہ ر
199
Verbs
ں
/ham gaē hōṁ gē/ we must/will have
1st person plural
ں /ham gaī hoṁ gī/ gone
2nd person plural ں آپ /āp gaē hōṁ gē/ you must/will have
(formal) ں آپ /āp gaī hōṁ gī/ gone
Formation:
perfective participle + /hōnā/ (imperfective participle)
The past irrealis or perfective irrealis describes contrary-to-fact events which, had they occurred,
would have been completed in the past.
Table 8.27 shows the conjugation of the perfective irrealis form of the verb /jānā/ to go.
200
Verbs
201
Verbs
As a marked, masculine noun, the infinitive may be inflected and used in any position in the
sentence, in the direct or oblique case, and may take postpositions and compound postpositions,
just like any other noun. In the direct case, it can be the subject of a sentence or occur in indirect
constructions, and in the oblique case, in constructions with either postpositions or main verbs.
The infinitive is used in the direct case with the following verbs. Optionally, these nominative
infinitives may be inflected as adjectives, to agree with the nouns they govern.
• /cāhnā/ to want
(8.77) ں ب بڑ
mæṁ paṛh-n-ā cah-t-ī hūṁ
I.DIR read-INF-M.SG.DIR want-IP-F.SG be.PRS.1.SG
‘I want to study.’
• /sīkʰnā/ to learn
Uninflected infinitives may be used for requests and instructions where the connotation is one
of neutrality on the formality scale. This usage is most common in impersonal contexts, such as
directions and recipes. One term for this is the distanced imperative. That is, it is used to give
instructions, directions, orders which are to be carried out at a distance, either of time or space.
202
Verbs
bʰāg kē dikʰ-ā-n-ā
run CP see-CAUS-INF-DIR
‘Show me how you run!’
8.6.3.1 As Subject
Some common indirect constructions are built around infinitives in the direct case.
The indirect construction is a sentence type in which the verb agrees, not with the doer or expe-
riencer of the action--the logical subject of the sentence--but with the logical direct object. The
noun referring to the doer/experiencer of the action--if it appears in the sentence at all--is marked
with a postposition, usually @ /kō/, or with the alternate form of the personal pronoun--/mujʰe,
203
Verbs
tujʰe/ and so on. (See Table 4.2 for more on those alternate forms and Section 5.1.2.3 for more
on indirect constructions.)
Infinitives used in the constructions below may take objects of their own. When they do, these
infinitives generally take adjectival inflection and agree with their objects, though the prevalence
of such agreement varies both with the length of the sentence (the longer the sentence, the more
optional the agreement) and with the dialect (in some dialects the infinitive remains masculine
singular). The main verb, however, must agree with the object of the infinitive, unless the object
is followed by a postposition.
Negative indirect constructions are formed with the particle ب/nahīṁ /, unless they are in the
subjunctive or occur in an irrealis sentence; these two latter cases use the negative particle بہ/nah/.
In Pakistani Urdu, either @ /kō/ or /nē/ may follow the logical subject of the sentence; usually,
@ /kō/ explicitly signifies lack of control of the circumstances or event on the subject’s part, due
to external obligation or compulsion, while /nē/ covers all other cases, either where there is
subject volition or where the circumstances are neutral.
204
Verbs
(8.86) ؟
tum nē jā-nā
you ERG go-INF
‘Are you going to go? (+ conscious choice)’
versus
(8.87) @ ،ں
In the past tense, /cāhiē/ may be followed by a past tense inflected form of /hōnā/, which
then becomes the main verb and agrees in number and gender with the grammatical subject.
This construction means to know (how); it is used for learned skills or behaviors.
21
In origin, a frozen passive form meaning is wished, is necessary.
205
Verbs
(8.90) ب a
• with postpositions
• with verbs
206
Verbs
When the action of the verb has not yet occurred, this construction denotes imminence, and may
be translated with the phrase about to; this usage is comparable in meaning to the construction
below, with @ /kō/.
(8.97) @ ا ن
imtihān hō-n-ē kō hæ
exam be-INF-OBL ACC be.PRS.3.SG
‘The exam is starting soon.’
23
See Section 5.1.2.
207
Verbs
This construction may also be used to show purpose. (See below, Section 8.6.4.4).
(8.98) a ب ز ن ر
zīšān mār kʰā-n-ē mēṁ māhir hæ
Zishan beat eat-INF-OBL in skilled.person be.PRS.3.SG
‘Zishan is skilled at getting beaten up.’
(8.100) ڈو
kištī ḍūb-n-ē lag-ī
boat sink-INF-OBL attach-PP.F.SG
‘The boat began to sink.’
(8.101) رش
šāyad bāriš hō-n-ē lag-ē
maybe rain be-INF-OBL attach-SBJV.3.SG
‘It may begin to rain.’
It may be used with stative verbs to describe conditions that began in the past but continue in
the present; this usage is most common with perfective tenses.
(8.102) رش
bāriš hō-n-ē lag-ī hæ
rain be-INF-OBL attach-PP.F.SG be.PRS.3.SG
‘It has begun to rain.’
24
See Section 5.2.3.
25
See Section 5.2.1.
208
Verbs
• Infinitive[OBL] + @ /kō/
(8.104) @ @ ھ
bahot kucʰ kʰā-n-ē kō jī cāh-t-ā hæ pī-n-ē
very some eat-INF-OBL ACC heart want-IP-M.SG be.PRS.3.SG drink-INF-OBL
kō jī cah-t-ā hæ
ACC heart want-IP-M.SG be.PRS.3.SG
‘I feel like eating a lot of things and drinking a lot of things.’
209
Verbs
Strong negative assertion may be expressed with ب /nahīm/, an oblique infinitive, and the
particle ، ، 5 /kā, kē, kī/ acting as the main verb of the sentence, and agreeing with the subject.
In addition to /hōnā/, Urdu has six other auxiliary verbs that can be combined with a verb stem
or participle to express a category of mood or aspect. The mood of a verb has to do with either (1)
the speaker’s view of the reality of the action or event to which the verb refers (for example, that
it could happen, should happen, is contingent on something, etc); or (2) the intent of the speaker
(as with a command, where the speaker intends that the addressee perform the expressed action;
this is often referred to as the imperative mood).26 Aspect refers to the structure of the action or
event with respect to time--not when it happens (its tense), but its temporal character: whether it
is completed, ongoing, repeated, and so on. The modal or aspectual meaning that each auxiliary
adds is predictable and well-defined; this predictability is a chief difference between auxiliary
verb constructions and vector verb constructions, which are discussed in the next chapter.
In the first six of the auxiliary constructions below, the auxiliary verb, which is intransitive, deter-
mines the transitivity of the entire verbal form; therefore the ergative marker /nē/ is not used
with any of them, even when the other verb, which carries the meaning, is transitive. The seventh
construction, which uses : /karnā/, is not used in perfective tenses, so the issue of transitivity
does not arise with it.
This construction expresses the ability of a subject to perform an action or the possibility that a
given event will occur. /saknā/ occurs only as an auxiliary verb, never by itself.
26
The reason for the term mood is that both (1) and (2) have to do with the speaker’s attitude toward the action.
210
Verbs
Formation:
verb stem + /saknā/ to be able
/pānā/ occurs most frequently in negative sentences, expressing contingent action or contingent
ability. Note that when it occurs as an independent verb, it is transitive.
Formation:
verb stem + /pānā/ to find, to obtain, to receive
211
Verbs
This construction indicates that an event or action has taken place--and been completed--before
another event or action. It is most often used in perfective tenses.
Formation:
verb stem + /cuknā/ to be finished
This construction occurs with stative verbs; that is, verbs that denote a state rather than an action,
such as the English verbs to be, to believe, to seem, as opposed to verbs like to hit, to bounce, to climb.
Hence it expresses the continuation of a state or an activity that can be progressive or ongoing,
rather than one that can be stopped and started over and over. (Compare the meaning of the
following /rahnā/ construction in Section 8.7.5, which involves non-stative verbs, representing
activities that can be repeated, or iterative, rather than progressive, or continuous.)
Formation:
perfective participle + ر/rahnā/ to remain, to stay
212
Verbs
In contrast with the previous construction, this one occurs with dynamic, or non-stative, verbs
and emphasizes the continuity or repeated nature of an activity. It is not used in the negative nor
in progressive tenses.
Formation:
imperfective participle + ر/rahnā/ to remain, to stay
This construction can express an ongoing activity that is deliberate or that will result in change.
213
Verbs
Formation:
imperfective participle + /jānā/ to go
Deliberate:
Resulting in change:
(8.121) ؤ @
This construction, which consists of the masculine singular perfective participle of : /karnā/ +
a non-perfective tense expresses an action that is repeated at intervals which are not regular or
predictable. As noted in the introduction above, it does not occur in the perfective tenses. The
distinction between the past imperfect (see Section 8.4.2.2) and this construction with : /karnā/
in the imperfect is being lost for some speakers.
Formation:
perfective participle + : /karnā/ to do
214
Verbs
This is one of only two constructions (the other is the passive) in which a regularly formed per-
fective participle of /jānā/ ( /jāyā/) is used, rather than the more common گ/gayā/. See
the following example.
215
Verbs
</Mo:InflectionalAffix>
216
Verbs
</Mo:InflectionalAffix>
217
Verbs
218
Verbs
<Ph:refPhoneme idref="phLongI"/>
</Ph:SimpleContextTerminal>
</Ph:leftContext>
</Ph:Environment>
</Mo:AffixAllomorph>
<Mo:AffixAllomorph>
<!-- -īm
̇ -->
<Ph:Form spelling="+ "ب/>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvParticiple'/>
<Fs:SymbolicFeatureValue defn='fvPerfective'/>
<Fs:SymbolicFeatureValue defn='fvFeminine'/>
<Fs:SymbolicFeatureValue defn='fvPlural'/>
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
219
Verbs
220
Verbs
<Mo:AffixAllomorph>
<!-- -ūm
̇ -->
<Ph:Form spelling="+"وں/>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvFinite'/>
<Fs:SymbolicFeatureValue defn='fvSubjunctive'/>
<Fs:SymbolicFeatureValue defn='fv1'/>
<Fs:SymbolicFeatureValue defn='fvSingular'/>
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
<Mo:InflectionalAffix gloss="-Subjunctive.2Sg" id="afSubjunctive2Sg">
<Mo:AffixAllomorph>
<!-- -ē -->
<Ph:Form spelling="+"ے/>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvFinite'/>
<Fs:SymbolicFeatureValue defn='fvSubjunctive'/>
<Fs:SymbolicFeatureValue defn='fv2'/>
<Fs:SymbolicFeatureValue defn='fvSingular'/>
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
<Mo:InflectionalAffix gloss="-Subjunctive.3Sg" id="afSubjunctive3Sg">
<Mo:AffixAllomorph>
<!-- -ē (homophonous with 2sg)-->
<Ph:Form spelling="+"ے/>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvFinite'/>
<Fs:SymbolicFeatureValue defn='fvSubjunctive'/>
<Fs:SymbolicFeatureValue defn='fv3'/>
<Fs:SymbolicFeatureValue defn='fvSingular'/>
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
221
Verbs
</Mo:InflectionalAffix>
<Mo:InflectionalAffix gloss="-Subjunctive.2InfPl" id="afSubjunctive2InfPl">
<Mo:AffixAllomorph>
<!-- -ō (homophonous with mid-level imperative)-->
<Ph:Form spelling="+"و/>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvFinite'/>
<Fs:SymbolicFeatureValue defn='fvSubjunctive'/>
<Fs:SymbolicFeatureValue defn='fv2'/>
<Fs:SymbolicFeatureValue defn='fvInformal'/>
<Fs:SymbolicFeatureValue defn='fvPlural'/>
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
<Mo:InflectionalAffix gloss="-Subjunctive.2FormPl" id="afSubjunctive2FormPl">
<Mo:AffixAllomorph>
<!-- -ēm (homophonous with 1 plural) -->
<Ph:Form spelling="+ "ب/>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvFinite'/>
<Fs:SymbolicFeatureValue defn='fvSubjunctive'/>
<Fs:SymbolicFeatureValue defn='fv2'/>
<Fs:SymbolicFeatureValue defn='fvFormal'/>
<Fs:SymbolicFeatureValue defn='fvPlural'/>
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
<Mo:InflectionalAffix gloss="-Subjunctive.3Pl" id="afSubjunctive3Pl">
<Mo:AffixAllomorph>
<!-- -ēm (homophonous with 1 plural) -->
<Ph:Form spelling="+ "ب/>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvFinite'/>
<Fs:SymbolicFeatureValue defn='fvSubjunctive'/>
<Fs:SymbolicFeatureValue defn='fv3'/>
<Fs:SymbolicFeatureValue defn='fvPlural'/>
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
<Ph:Environment id="envAfterLongV">
<Ph:leftContext>
<Ph:AlternativesContext>
222
Verbs
<Ph:Environment id="envNotAfterLongV">
<Ph:leftContext>
<Ph:AlternativesContext>
<!--The negation of a natural class (better thought of as
the elsewhere case); but since the system does not allow
for negated classes, we define it as the conjunction of
everything that is not a long vowel. Note that some of
these phonemes (like the short vowels) are possible, but
unlikely, in the Arabic script.-->
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phShortA'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phShortI'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phShortU'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phShortE'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phShortO'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phEpsilon'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phOpenO'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
223
Verbs
<Ph:refPhoneme idref='phZabar'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phZer'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phPesh'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phNunGunna'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phHamza'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phSukun'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phTashdid'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phAlif'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phBe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phBhe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phPe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phPhe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phTe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phThe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phTDotE'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phThDotE'/>
</Ph:SimpleContextTerminal>
224
Verbs
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phSe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phJim'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phJhe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phCe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phChe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phBarihe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phXe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phDal'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phDHe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phDDotal'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phDDotHe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phZal'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phRe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phRDotE'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phRhe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phZe'/>
225
Verbs
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phZhe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phSin'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phShin'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phSvad'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phZvad'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phToe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phZoe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phAyin'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phGain'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phFe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phQaf'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phKaf'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phKhe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phGaf'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phGhe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
226
Verbs
<Ph:refPhoneme idref='phLam'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phMim'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phNun'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phVau'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phChotiHe'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phYeh'/>
</Ph:SimpleContextTerminal>
<Ph:SimpleContextTerminal>
<Ph:refPhoneme idref='phDoCashmiHe'/>
</Ph:SimpleContextTerminal>
</Ph:AlternativesContext>
</Ph:leftContext>
</Ph:Environment>
The formal grammar template for verbal morphology is given below; note that because the formal
grammar is limited to morphology, it does not include the various synthetic forms (i.e. the forms
using auxiliary verbs).
227
Verbs
<Mo:refInflectionalAffix idref="afPfctvPtcplFmPl"/>
<!--Infinitives-->
<Mo:refInflectionalAffix idref="afInfDirect"/>
<Mo:refInflectionalAffix idref="afInfOblique"/>
</Mo:InflAffixSlot>
</Gr:PartOfSpeech>
228
Complex Predicates
Chapter 9
Complex Predicates
9.1 Introduction
In Urdu, complex verbal forms can be created by combining an inflected verb with another word--
either a noun or adjective, or a non-finite (non-inflected) verb--in the pattern X + Inflected Verb.
These two kinds of complex predicates are categorized according to the identity of X:
The first of these constructions makes a verbal form out of a noun or adjective. The second
construction creates a new verb by adding a second, “vector” verb to the main verb, which adds
to the main verb’s meaning in a way that is somewhat generalizable but can differ with context.
Combining the two verbs sometimes even creates a new lexical item with a meaning connected
to, but not predictable from, its components. Both types of construction are also known as light
verb constructions, because the second element, the inflected verb, loses its literal meaning (and
hence becomes “light”) when combining with X.
The difference between vector verb and auxiliary verb constructions (Section 8.7) is that auxiliary
verbs add a predictable nuance--either aspectual or modal--to the first verb, whose core meaning
does not change, whereas in vector verb constructions, the combination of the two verbs is less
predictable.
As we said above, denominative constructions are a way of turning an adjective or noun into a
verb;3 for example, م5 /kām/ work (n.) + : /karnā/ to do → to work; 5 د/dʰākā/ push (n.) +
1
Also called conjunct verbs by some linguists.
2
These are called compound verbs by many South Asianist linguists.
3
The term denominative is from Latin and loosely means from a noun.
229
Complex Predicates
6 د/dēnā/ to give → to push (someone). Denominatives also occur frequently with loanwords as
a means of borrowing verbs without violating the pan-South Asian intolerance of directly bor-
rowing verbs as verbs. Verbal loanwords (from Arabic, participles or verbal nouns, and from
Persian, present or past stems) are paired with an inflected Urdu verb to produce a new verb
form. Borrowed nouns and adjectives can also be used, and English verbs are borrowed in this
way as well:
Denominatives from English loanwords:
(9.1) :واہ! ر@ ا ن س
230
Complex Predicates
: /karnā/ and /hōnā/ are the most common verbs in denominative constructions and play
a special role in that capacity, forming parallel transitive-intransitive pairs. The same noun or
adjective can be paired with : /karnā/ to make a transitive form and with /honā/ to make
an intransitive; for example, : دا/dāxil karnā/ to enroll and دا/dāxil hōnā/ to enter, to be
enrolled.
Other verbs that can be used in denominative constructions include 6 د/dēnā/ to give, /lagnā/
to be applied, be put, be joined (to), /lēnā/ to take, /bāṁ dʰnā/ to tie, /nikālnā/ to take
out, /lagānā/ to apply, ا/ut ̣ʰānā/ to raise, and ر/rakʰnā/ to put. Note that all of these verbs
are transitive, as are the denominative constructions they form.
9.2.2 Examples
231
Complex Predicates
(9.9) @ ب آ آپآوازد
Each of the vector verbs below can be combined with a variety of verbs to add extra details or
nuance to the core meaning of that verb, or to form an entirely new, compound lexical item. This
latter case is called relexicalization:through the combination of the two verbs, a new word is cre-
ated, whose meaning is related to, but not predictable from, the meanings of its components. The
examples below will illustrate the different ways vector verbs can add to or change the meaning
of the main verb.
Some vector verbs are intransitive and some transitive. See below for more on transitivity.
Formation:
verb stem + inflected vector verb4
9.3.1.1 /jānā/ to go
Used with verbs of motion and stative verbs, but only ones that represent completable actions,
/jānā/ expresses a change of state through the action of the main verb.
4
See following sections for lists of possible vector verbs.
232
Complex Predicates
(9.10) @وہآ ۓ
As a vector verb, بڑ/paṛnā/ is used only with non-stative verbs and expresses sudden, unexpected,
involuntary, or unpreventable action.
The vector verb /nikalnā/ expresses sudden or unexpected action, as well as motion out or
away.
(9.13) :واہ! ر@ ا ن س
233
Complex Predicates
ا/ut ̣ʰnā/ works as an intensifier of the main verb and also expresses sudden action. Even when
this construction contains a transitive main verb, that verb cannot take an object. (See the third
bullet in Section 9.3.3.)
akbar k-ī mɔt k-ī xabar sun kar abrār cīx ut ̣ʰ-ā
Akbar POSS-F.SG death POSS-F.SG news hear CP Abrar yell rise-PP.M.SG
‘Having heard the news of Akbar’s death, Abrar began screaming.’
/bæt ̣ʰnā/ can suggest a sense of impulsiveness or involuntary quality to the action, or express
that the action is an unrectifiable mistake. This vector verb may also overlap with the category of
mood in that it sometimes implies speaker disapproval of the action. Unlike the previous vector
verbs, it is used most often with transitive verbs, which may still take objects. The construction
as a whole, however, behaves like an intransitive verb in the perfective tenses.
234
Complex Predicates
In keeping with its semantics, 6 د/dēnā/ expresses an action that is done on behalf of someone
else, or is directed away from oneself, or in some other way has an external effect. It can also
signify completion of the action, and in requests, adds a slight sense of politeness.
In parallel with the outward semantics of 6 د/dēnā/, /lēnā/ expresses an action that one per-
forms for or towards oneself or that has an external origin but affects the self. Like 6 د/dēnā/, it
can also signify completion of the action.
235
Complex Predicates
ڈا/ḍālnā/ can intensify the main verb, adding a sense of urgency or violence, as well as express
completeness.
ر/rakʰnā/ occurs with a limited number of verbs, chiefly 6 د/dēnā/ and /lēnā/ and refers to
the continuity of a particular state.
236
Complex Predicates
Intransitive vectors are generally used with intransitive main verbs, and transitive vectors with
transitive main verbs, although there can be exceptions to this generalization in either direction
(intransitive main verb with transitive vector verb or transitive main verb with intransitive vector
verb). The transitivity of the sentence is determined according to the list below. In an intransitive
sentence, the vector verb agrees with the subject even in perfect tenses and the subject does not
take /nē/; whereas in a transitive sentence, the vector verb agrees with the object in the perfect
tenses while the subject takes /nē/:
237
Complex Predicates
(9.32) وہ ر رف د
voh gʰar k-ī taraf cal dī-yē
They house POSS-F.SG side walk give-PP.M.PL
‘They went off towards home.’
There are some types of verb constructions that either cannot occur with vector verbs or else do
so only rarely:
• auxiliary constructions with /saknā/, except (rarely) with relexicalized compounds (see Sec-
tion 8.7.1)
238
Complex Predicates
239
The Formal Grammar
Appendix A
This section contains the skeleton of the formal grammar, with the various grammar fragments
defined elsewhere linked in by the use of XML tags. These tags appear as labels in this display, but
consist of 'src:fragref' tags in the underlying XML code. The id 'top' on the top-most 'src:fragment'
tag indicates to the XSLT style sheet responsible for extracting the formal grammar that this is the
top-most piece of the complete formal grammar, i.e. the one under which all the other fragments
are to be embedded when the formal grammar is extracted.
<!--=================Morphological Data=====================-->
240
The Formal Grammar
<Mo:morphologicalData>
<!--==============Strata=================-->
<!--We do not define any Strata-->
<!--============Derivational Affixes==========-->
<!--We do not define any derivational affixes====-->
<!--============Inflectional Affixes==========-->
<!--===========Noun suffixes:============-->
<src:fragref linkend="fragNounSuffixes">
<!--===========Verb suffixes:============-->
<src:fragref linkend="fragVerbSuffixes">
<!--===========Adjective suffixes:============-->
<src:fragref linkend="fragAdjectiveSuffixes">
</Mo:morphologicalData>
<!--=================Phonological Data=====================-->
<Ph:phonologicalData>
<!--Phoneme Sets-->
<src:fragref linkend="fragUnicodePhonemes">
<!--Environments-->
<src:fragref linkend="fragVerbEnvironments">
<!--Phonological Rules-->
<src:fragref linkend="fragSpellingRules">
</Ph:phonologicalData>
<!--================Irregular Words=====================-->
<Irr:lexicalData>
<!--Verb honaa -->
<src:fragref linkend="fragIrregularHona">
</Irr:lexicalData>
</Ln:LanguageData>
241
The Grammar for Dictionary Importation
Appendix B
This section contains the skeleton of a separate grammar whose sole purpose is to convert the
citation forms of words into their stems. For nouns, this task includes inferring the declension class
from the phonological shape of the noun (particularly its ending) and grammatical information
(specifically, its gender).
242
The Grammar for Dictionary Importation
xmlns(Fs=http://casl.umd.edu/LinguisticSchemas/FeatureSystem.rng)
xpointer(//src:fragment/node())"
/>
</Fs:FeatureSystem>
<Mo:morphologicalData>
<!--We import all the suffixes, but only the ones referenced by the
template (above) are actually used.-->
<xi:include href="fragAdjectiveSuffixes.xml"
xmlns:xi="http://www.w3.org/2001/XInclude"
xpointer='xpointer(id("fragAdjectiveSuffixes")/node())'/>
</Mo:morphologicalData>
<Ph:phonologicalData>
<!--Import phonological information that is used by two or more of
the *.import.xml files: -->
<xi:include href="import.phonology.xml"
xmlns:xi="http://www.w3.org/2001/XInclude"
xpointer="xmlns(src=http://casl.umd.edu/LiterateProgramming)
xmlns(Ph=http://casl.umd.edu/LinguisticSchemas/Phonology.rng)
243
The Grammar for Dictionary Importation
xpointer(//src:fragment/node())"
/>
</Ph:phonologicalData>
</Ln:LanguageData>
244
The Grammar for Dictionary Importation
<!--While many suffixes are listed for this slot in the main grammar,
only one concerns us for purposes of dictionary import: -->
<Mo:refInflectionalAffix idref="afInfDirect"/>
</Mo:InflAffixSlot>
<Mo:morphologicalData>
<!--This contains the single verbal suffix in the main grammar that is
used to mark the citation form of verbs in Urdu dictionaries.-->
<Mo:InflectionalAffix gloss="-Inf.Direct" id="afInfDirect">
<Mo:AffixAllomorph>
<Ph:Form spelling="+ "/>
</Mo:AffixAllomorph>
<!--No need to worry about features-->
</Mo:InflectionalAffix>
</Mo:morphologicalData>
<Ph:phonologicalData>
<!--Import phonological information that is used by two or more of
the *.import.xml files: -->
<xi:include href="import.phonology.xml"
xmlns:xi="http://www.w3.org/2001/XInclude"
xpointer='xpointer(id("import.phonology.xml")/node())'
/>
</Ph:phonologicalData>
</Ln:LanguageData>
245
The Grammar for Dictionary Importation
246
The Grammar for Dictionary Importation
</Mo:InflAffixSlot>
</Gr:PartOfSpeech>
<Mo:morphologicalData>
<!--This contains the subset of the noun affixes in the main grammar's
list which may appear in citation forms, namely:
the masculine plural direct suffix (for pluralia tantum words)
the feminine plural direct suffix
the masculine singular direct suffix
the feminine singular suffix
The affixes are listed in the above order, with the plurals before
the singulars, so that the plurals (which are marked at import time
by both the relevant gender fature and the plural value of the number
feature) take precedence over the singulars (which are marked for
gender at import, but not for the default singular value). (The order
of masculine vs. feminine is not important, since both genders are
marked at import.)
In a few cases, the main grammar says that a plural direct suffix
allomorph can appear with > 1 inflection class. For example, in the
feminine, the plural suffix -īyām
̇ can appear with either feminine
marked nouns ending in the singular with either /-ī/ or /-īyā/.
These are different declension classes, and in theory if the citation
form ends in -īyām, we couldn't tell which of these two declension
classes the word was. But in fact, these two declension classes
are identical in all cases of the plural; and since the only reason
for giving a plural citation form is that the noun only occurs in
the plural, we don't need to distinguish these two declension classes
(and we arbitrarily choose one for such pluralia tantum words).
-->
<Mo:InflectionalAffix gloss="-MascPlDirect" id="afMascPlDirect">
<Mo:AffixAllomorph>
<!-- -ae for class Y (class Y also allows -ē, but that suffix
247
The Grammar for Dictionary Importation
248
The Grammar for Dictionary Importation
249
The Grammar for Dictionary Importation
<Ph:Form spelling="+"ا/>
<Mo:inflectionClasses>
<Mo:refInflClass idref="classNounMarkedV"/>
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Mo:AffixAllomorph>
<!--Null allomorph-->
<Ph:Form spelling="+"/>
<Mo:inflectionClasses>
<Mo:refInflClass idref="classNounUnmarked"/>
</Mo:inflectionClasses>
</Mo:AffixAllomorph>
<Fs:FeatureStructure description='Inflectional Features'>
<Fs:SymbolicFeatureValue defn='fvMasculine'/>
<Fs:SymbolicFeatureValue defn='fvSingular'/>
<Fs:SymbolicFeatureValue defn='fvDirect'/>
</Fs:FeatureStructure>
</Mo:InflectionalAffix>
250
The Grammar for Dictionary Importation
</Mo:morphologicalData>
<Ph:phonologicalData>
<!--Import phonological information that is used by two or more of
the *.import.xml files: -->
<xi:include href="import.phonology.xml"
xmlns:xi="http://www.w3.org/2001/XInclude"
xpointer='xpointer(id("import.phonology.xml")/node())'
/>
</Ph:phonologicalData>
</Ln:LanguageData>
251
References Cited or Consulted
Barker, Muhammad Abd-Al-Rahman, Hasan Jahangir Hamdani, Khwaja Muhammad Shafi Dihlavi
& Shafiqur Rahman. 1967. Spoken Urdu: A course in Urdu, vol. I-III. Montreal: Institute of
Islamic Studies, McGill University.
Barz, Richard & Yogendra Yadav. 2000. An introduction to Hindi and Urdu. Munshiram Manoharlal
Publishers Pvt. Ltd.
Bashir, Elena. 1999. The Urdu postposition ne: its changing role in the grammar. In Rajendra
Singh (ed.), The Yearbook of South Asian Languages and Linguistics 1999. New Delhi, SAGE
Publications Pvt. Ltd.
Unicode Consortium, The. 2007. The Unicode standard, version 5.0. Addison-Wesley Professional,
5th edn.
Davis, Mark & Martin Dürst. 2008. Unicode normalization forms. Unicode standard annex #15.
URL http://unicode.org/reports/tr15/.
Knuth, Donald E. 1992. Literate programming, CSLI Lecture Notes. Stanford: Center for the Study
of Language and Information.
Matthews, David & Mohamed Kasim Dalvi. 2003. Teach yourself Urdu. McGraw-Hill Companies
Inc.
Maxwell, Michael. 2006. Final technical report M.1: Current morphological methods. Tech. rep.,
CASL, University of Maryland, College Park, MD.
Maxwell, Michael. 2007. Final technical report M.5: Standards for lexical and morphological
interchange.
Naim, C.M. 2000. Introductory Urdu, vol. I. New Delhi, National Council for Promotion of Urdu
Language.
Pray, Bruce R. 1970. Topics in Hindi-Urdu grammar. Berkeley, Center for South and Southeast Asia
Studies.
Qureshi, Bashir Ahmad. 2003. Standard Twenty-First century dictionary: Urdu into English. Revised
and enlarged by Abdul Haq. Delhi, Educational Publishing House.
252
References Cited or Consulted
Walsh, Norman. 2002. Literate programming in XML. In XML 2002. Baltimore, MD. URL http:
//nwalsh.com/docs/articles/xml2002/paper.pdf.
Walsh, Norman & Leonard Muellner. 1999. DocBook: the definitive guide. Sebastopol, California:
O'Reilly & Associates, Inc. URL http://www.docbook.org/.
253
Index
Index
/al/, see Affixes, /al/, Arabic definite article repetition of, 148
superlative
Abbreviations, 35 in /sab sē/, 118
Ablative, see Case, ablative in /tarīn/, 118
Adjectives unmarked, 110
/-ṁ / inflection loanwords, 113
inflection of, 109 y-v-k-j class
with ordinal numbers, 115 as adverbs, 123
agreement, 105, 109 inflection of, 116
attributive, 113 y-v-k-j pronouns as, 75, 78
comparative, 117 Adverbs
as intensifiers, 117 /kahīṁ /
in /sē/, 97, 118 in vector verb constructions, 238
in /tar, tarīn/, 117 agreement, 121
in /zyāda/, 118 complex
demonstrative, 116 of time, 121
in subordinate constructions, 138 compound, see Adverbs, phrasal
derivation of in /is/, /us/, /kis/, /jis/, 121, 123
from nouns, 111, 113 demonstrative, 121
from verbs, 113 pronouns as, 121
in loanwords, 111 derivation of
with /ham/, 62 from oblique nouns and adjectives, 116
imperfective participles as, 192 emphatic particles as, 128
in complex verbs, 229 interrogative, 121
infinitives as, 202, 204 vs relatives, 123
interrogative, 116 loanwords, 124
marked, 105 of manner, 123, 124
masculine and feminine marked of place, 123, 124
inflection of, 105 of time, 123
negative, 113 phrasal, see Adverbs, compound, 124
oblique with /sē/, 124
as adverbs, 121 relative, 121
with /kā/, 88 replaced by /keh/, 136
phrasal, 118 suffixed with /bʰī/, 131
with /kā/, 83 vs interrogatives, 123
with /sā/, 118 with /hī/, 127
possessive, see Pronouns, possessive repetition of, 149
predicative, 113 with /kā/, 89
relative, 116
254
Index
255
Index
256
Index
257
Index
258
Index
259
Index
260
Index
261
Index
262
Index
/ɔr/ intangible, 85
as a qualifier, 131, 132 with /kē pās/, 85
coordinating conjunction, 132 with /kō/
in compounds, 131 intangible, 94
conjunctions, see Conjunctions Postpositions, 82
interjections, 145 /kā/, /kē/, /kī/, 82
negative, see Negative particles as object marker, 92
in commands, 144 in compounds, 85–91, 193
in subjunctives, 144 in possessive constructions, 83
sentential, 142 inflection of, 82
Perception versus enclitic /-e-/ (ezafe), 62
verbs of with adverb, 89
with /keh/, 137 with oblique adjective, 88
Plural with oblique noun, 87
agreement with oblique noun + postposition, 85
with /yeh/, 71 with Perso-Arabic preposition + oblique
auxiliary marking noun, 90
shifted to participle, 155 /kē liyē/
disambiguated by /lōg/, 71 with oblique infinitives, 209
in interrogatives, 78 /kō/
expressed through repetition as logical subject marker, 93
of nouns, 146 as object marker, 92
of pronouns, 147 in indirect constructions, 93, 203
oblique in passive sentences, 193
of cardinal numbers, 115 in perfect tenses, 196
of loanwords in possessive constructions, 94
Arabic, 64, 65 omission of, 204
Arabic and Persian, 58 versus /nē/ in infinitive constructions,
English, 58 95
Persian, 58 with oblique infinitives, 207, 209
of singular referents, 153 with pronouns, 71
with /ham/, 71 /mēm/̇
with /tum/ and /āp/, 70 as locative marker, 99
of Urdu nouns, see Nouns, masculine Or /mēṁ /
feminine, inflection of as locative marker, 99
verbs with /tum/ and /āp/, 70 of change of state, 100
with indefinite pronouns, 80 of cost, 100
“broken” plurals, 65 with infinitives, 101
Possession /mēṁ /̇
alienable, 85 with oblique infinitives, 208
in compounds /nē/, see Particles, 95
with enclitic /-e-/ (ezafe), 61 as volitional marker, 95
inalienable, 84 in infinitive constructions, 95
intangible, 85 in perfect tenses, 196
with /kō/, 94 in perfective tenses, 209
with /kā/, 83 versus /kō/ in infinitive constructions,
inalienable, 84 95
263
Index
264
Index
265
Index
266
Index
y-v-k-j class
adjectives, 116
as adverbs, 116
as nouns, 116
inflection of, 116
adverbs, 121
inflection of, 121
repetition of, 149
pronouns, 75–79
267