The Computational Morphology of Pashto Nouns

SOUTH ASIAN LANGUAGE REVIEW
VOL.XVII. No. 1, January 2007.

The Computational Morphology of Pashto Nouns
Mohammad Abid Khan and Fatima Tuz Zuhra
University of Peshawar, Pakistan

Abstract. This paper investigates the inflectional properties of Pashto
nouns. The main focus is on the classification of the Pashto nouns in
such a way that further computational work on them becomes easy. The
works of three leading Pashto grammarians, Penzl (1955: 51-65),
Reshteen (1994: 78-91), and Zyar (2003: 44-46) provide the basis for the
proposed classification. The characteristics of nouns, having a particular
type of ending phoneme, are closely observed from a large collection of
examples. Consequently, the Pashto nouns are classified into seven
classes for masculine nouns and seven classes for feminine nouns. The
class membership of each noun is unambiguous and the morphological
rules for these classes are uniform. These uniform rules have been
sketched with the help of Finite State Transducers (FSTs) ready for
implementation. The implementation of these FSTs will result in part of
a morphological analyzer.

1. Introduction
A Pashto noun refers to a person, an animal, an object or thing, a substance, a
state, an event, a quality, an abstract idea or some kind of feeling. An example,
of each of this category respectively, is saɻe ‘man’, ɣwɑ ‘cow’, korsəi ‘chair’,

obə ‘water’, wiara ‘fear’, bɑrɑn ‘rain’, neeki ‘goodness’, dastur ‘custom’ and
xolus ‘sincerity’.
A Pashto noun can be a subject or an object of a verb. It can be substituted
by pronouns. It can be modified by an adjective. It can have an article or
determiner before it. Consider example (1).
(1)
haɣә
yәw
Xɑista
as
wɑxeest-o.
he-OBL a
beautiful
horse bought(3sg, mas)
‘He bought a beautiful horse.’
In example (1), the pronoun haɣə ‘he’ is the subject and as ‘horse’ is the

object of the verb wɑxeeist-o ‘bought’; the noun as ‘horse’ is modified by the
adjective Xɑista...