Morphology and Syntax of the Tamil Language

Kengatharaiyer Sarveswaran
University of Jaffna. Sri Lanka.
[email protected]

Abstract
This paper provides an overview of the morphology and syntax of the Tamil language,
focusing on its contemporary usage. The paper also highlights the complexity and rich-
ness of Tamil in terms of its morphological and syntactic features, which will be useful
for linguists analysing the language and conducting comparative studies. In addition, the
paper will be useful for those developing computational resources for the Tamil language.
It is proven as a rule-based morphological analyser cum generator and a computational
grammar for Tamil have already been developed based on this paper. To enhance acces-
sibility for a broader audience, the analysis is conducted without relying on any specific
grammatical formalism.

Keywords: Tamil; Morphology; Syntax; Grammar; Dravidian; Complex morphology;

Computational morphology; Computational grammar.

1 The Tamil language

Tamil is a Southern Dravidian language spoken natively by more than 78 million people
across the world,1 including in India, Sri Lanka, Malaysia, Singapore, South Africa, Mau-
ritius, Fiji, and Burma (Krishnamurti, 2003). It is also one of the top 20 languages in
the world based on population count (Simons and Fennig, 2017). Tamil has more than
two millennia of continuous and unbroken literary tradition (Hart, 2000) and has been
recognised as a classical language by the Government of India. It has the longest literary
tradition among all the Dravidian languages (Lehmann, 1998). It is an official language of
Sri Lanka and Singapore and has official regional status in the states of Tamil Nadu and
Pondicherry in India. It has also been recognised as a minority and indigenous language
in several countries, including Malaysia, Mauritius, and South Africa. Tamil is taught in
schools in several countries, including Australia, Canada, and the United Kingdom.
The earliest classical Tamil is called sangattamil̲. Modern Tamil is known for its
diglossic nature whereby the spoken varieties, referred to as the low variety or kod̲untamil̲
and the written variety, also referred to as the high variety or centamil̲ (Britto, 1991;
1
http://www.languagesgulper.com/eng/Tamil.html

1
Krishnamurti, 2003), differ phonologically and morphologically. The spoken forms of
Tamil vary depending on the regions where they are spoken. This is mainly due to
language contact and politics (Schiffman, 2008). Within Sri Lanka, there are several
spoken varieties of Tamil. At times, one encounters cases where one speaker may not
understand the other. Further, no common agreement has been made among different
governments (at least between Sri Lanka and Tamil Nadu in India) on the choice of Tamil
words for certain technical terms. Consequently, terminological variation is also present.
Although there are variations in spoken Tamil, the linguistic structure of written Tamil
remains mostly uniform across the different regions.

1.1 Tamil script

The Tamil language has its own script, which is nowadays referred to as Tamil script
and follows the Abugida or Alphasyllabary writing system where a pure consonant and
a vowel are written together as one unit or syllable. For instance, the consonant த் (t)
together with the vowel உ (u) form the single unit, a composite character து (tu). The
alphabet has 12 vowels, 18 consonants (non-composite characters), and 216 composite
characters, which are formed when combining pure consonants and vowels together as a
single unit. Tamil also has a special character ஃ (ah) — a guttural, which is categorised
neither as a vowel nor as a consonant, but is in the alphabet. The University of Madras
lexicon describes this letter as “The 13th letter of the Tamil alphabet occurring only after
a short initial letter and before a hard consonant”. In addition to this total of 247 Tamil
letters, some letters from the Grantha alphabet, which is the script used to write Sanskrit
in South India, are also widely used along with the Tamil alphabet.
Apart from the alphabet, Tamil has its own native numerals, which include whole
numbers and fractions, metrics, and symbols for various concepts such as time that are
still being documented. It is, however, unfortunate that most current Tamil speakers
are not aware of these symbols, as they are not included in the school curriculum, even
though some of these symbols are still used in almanacs. However, they were commonly
used in books printed up until the early 20th century, with Tamil numerals even being
used for page numbering.

1.2 Encoding Tamil letters

Apart from the information about how the Tamil script works, understanding how it is
encoded in computers using Unicode2 is important in order to develop language processing
tools and to appreciate the complexities involved in processing the Tamil script. Each
syllable in Tamil has a single code point. For instance, all the vowels and pure consonants
2
https://home.unicode.org/

2
in the Tamil alphabet are each encoded with a single code point. All the composite
characters, on the other hand, are encoded with two Unicode points; one to represent
the consonant letter, and the other to represent the vowel modifier (also called dependent
vowel sign),3 which is applied to form the composite. For instance, து (tu) will have two
Unicode points,4 one for த (t) and the other for the ◌ு (u)- vowel modifier. Similarly,
ெதௗ (tou) also has two Unicode points corresponding to த (t) and ெ◌ௗ (ou) even though
it is represented by three glyphs. Typing these vowel modifiers and handling them as part
of computer programming is not a straightforward task. Unfortunately, some applications
still struggle with rendering these vowel modifiers.
The Tamil script was first encoded in the Unicode version 1.1 in the year 1993.5 Since
then, several Tamil language processing tools have been built around the world. Prior to
the introduction of Unicode, ASCII encoding was used along with special fonts to render
Tamil graphemes. Although this is not useful for Tamil language processing, this obsolete
approach is sometimes used to create text. This adds an extra level of challenges when
processing Tamil texts since these differences are not visible to the naked eye.

1.3 Tamil Grammars

The grammars of Tamil may be divided into ones that are composed by native scholars
and those that Europeans have written to facilitate the acquisition of the Tamil language,
mainly by foreigners (Pope, 1979). tolkāppiyam is recognised as the earliest scholarly work
on Tamil grammar by a native scholar (George, 2000). The date of this publication is not
exact, yet it is believed to have been published more than 2500 years ago. This is consid-
ered to be a derived work of an even older work called agastyam (Pope, 1979), the whole
work of agastyam is not extant. The other notable and still widely used and quoted Tamil
grammar is called nannūl, which is dated back to the 13th century. Modern-day Tamil
grammar textbooks used in Sri Lanka are themselves based on nannūl. Apart from tolkāp-
piyam and nannūl, several derived works have been published by native scholars from time
to time. An important factor about these works usually accommodates the evolution of
the Tamil language. Until the 16th century, these old grammars, dictionaries, treatises on
medicine, and literature were written in verse form, making them interpretable only by
Tamil pundits or experts — a non-trained person of the current generation would not be
able to understand this genre of Tamil. These texts are mostly understood through the
commentaries written by scholars. At times, even these commentaries require a further
simplified version in order to be understood. For instance, a scholar called parimēlaḻakar
wrote a commentary (Nachinarkkiniyar, 1937) for the first Tamil grammar work tolkāp-
3
https://unicode.org/charts/PDF/U0B80.pdf
4
https://unicode.org/charts/PDF/U0B80.pdf
5
https://www.unicode.org/standard/supported.html

3
piyam in the 13th century. Later, many simplified commentaries were published based
on parimēlaḻakar’s work, due to difficulties in understanding the structure and content of
those commentaries.
From the 16th century onwards, Europeans and others, mostly missionary scholars,
started publishing books in Tamil and writing Tamil grammar books. However, since
the 19th century, no notable complete linguistic studies have been done on Tamil (An-
namalai et al., 2014). On the other hand, the language has evolved significantly due to
modern subject areas, new kinds of literature, technological advancements, and influences
from other languages, particularly as a result of global communication and movement.
eḻuttu (orthography), col (morphology+syntax), pāeruḷ (semantics), yāppu (prosody), aṇi
(rhetorical embellishment) constitute the five key parts of Tamil grammar (Shanmugadas,
1982; Nachinarkkiniyar, 1937; Senavaraiyar, 1938). Not all of the grammars cover all of
these parts. For instance, the widely used Nannūl covers only the first two.
Tamil displays a relatively free constituent order, though it primarily follows a Subject-
Object-Verb (SOV) structure in formal writing. Using a corpus-based study, Futrell et al.
(2015) show that Tamil has the highest word order freedom when compared to the 100+
languages that are available in the Universal Dependencies treebank collection.
In traditional grammar, Tamil words are categorised into four types, namely nouns,
verbs, particles and intensifiers (Senavaraiyar, 1938; Thesikar, 1957). However, mod-
ern linguists classify Tamil words into peyar (noun), viṉai (verb), peyaraṭai (adjective),
viṉaiyaṭai (adverb) and iṭaiccol (particle) (Nuhman, 1999; Annamalai et al., 2014). Nomi-
native, accusative, dative, instrumental, sociative, locative, ablative, genitive and vocative
are the nine cases that mark nouns as described in the literature (Lehmann, 1993). In
addition to the case features, all nouns in Tamil can be categorised into uyartiṇai (ra-
tional) and aḥṟiṇai (irrational). Entities marked as rational are those perceived as being
able to think on their own, while the rest are termed as irrational. For instance, animals,
trees, furniture items, and infants are considered irrational, while humans and gods are
categorised as rational. This (ir)rationality based marking differs from splits in terms of
human vs. non-human, or animacy. For instance, infants are considered irrational just
as animals or inanimate objects, even though infants are human and animate. An adult
human entity can also still be marked as irrational if they behave in an insane manner.
This marking is in turn reflected in the noun-verb agreement. Tamil verbs have complex
morphosyntactic relations, which take auxiliary items such as question particles, emphatic
particles, and conjunctive markers, in addition to regular inflections such as tense, person,
gender, number, and honorifics to build a surface form.

4
2 Tamil Morphology
Tamil is an agglutinative language in which grammatical components are suffixed to root
words. These suffixes can be identified and separated from the rest of the word through
a complex segmentation process to determine the root words. This section provides an
overview of the morphology of nouns, verbs, adjectives, and adverbs.

2.1 Nominal Morphology

Nouns in Tamil are primarily marked for number and case. These features are expressed as
suffixes that are added to a 1. lemma form or 2. oblique form (which will be discussed later
in this section). In addition, an oblique form can take euphonic and other phonologically
motivated material before the number and case suffix (Caldwell, 1998; Shanmugadas,
1982; Lehmann, 1993).
The number or/and case suffix can be just added to a lemma form, as in example (1)
(a) and (b). Apart from such simple suffixation, a glide letter ( ய் / வ்) (y/v) may be
inserted, as in Example (1) (c), or a consonant (க்/ ப் / ச் / ற்) (k/p/c/ṟ) may be inserted,
as in Example (1) (d). In addition to these, an assimilation process may also occur where
an existing letter is changed to another, as in Example (1) (e), where ல் becomes ற்.

(1) (a) கதிைரகள் (katiraikaḷ) ‘chairs’

கதிைர -கள்
katirai -kaḷ
chair pl

(b) கதிைரக்கு (katiraikku) ‘to a chair’

கதிைர -க்கு
katirai -kku
chair dat

(c) கதிைரயுடன் (katiraiyudan) ‘with a chair’

கதிைர -ய் -உடன்
katirai -i -udan
chair glide soc

(d) பூக்கள் (pūkkaḷ) ‘flowers’

பூ -க் -கள்
pū -k -kaḷ
flower sandhi pl

(e) புற்கள் (puṟkaḷ) ‘grasses’

புல் -கள்
pul -kaḷ
grass pl

5
 Oblique forms are generated primarily by doubling the last consonant, adding a com-
pulsory oblique suffix, or by deleting last letter of the lemma. This process makes a stem
eligible to receive case markers in some noun classes (Krishnamurti, 2003). The com-
pulsory oblique suffixes are referred as cāriyai in Tamil grammar books (Nuhman, 1999;
Thesikar, 1957). அம் (am), அத்து (attu), and அற்று (aṭṭu) are the widely seen oblique
suffixes in texts. Examples for obliques are shown in (3).
Number and case suffixes are bound morphemes. However, in modern Tamil, case
suffixes can also be seen separately as free morphemes. Lehmann (1993) also points out
the possibility of their occurrence as free morphemes in Tamil.

2.1.1 Structure of simple Nouns

The morphology of a simple Tamil noun, without any derivations or compounding, is

depicted in the formulas in (2) (a) and (b). In Tamil text, nouns with two euphonic
markings can also be found, as shown in (2) (b). Euphonic morphemes such as அன் (an)
and இன் (in) are purely phonological increments (Lehmann, 1993). However, the exact
functions of these increments are yet to be fully understood. (3) illustrates various ways
a noun can be formed from an oblique base. A stem can take only one oblique suffix, but
it can accommodate multiple instances of euphonic morphemes.

(2) (a) Noun = lemma-form (+plural) (+case)

(b) Noun = oblique-form (+plural) (+euphonic)2 (+case)
(3) (a) மரங்களினால் (marangkaḷinaal) ‘by trees’
மரம் கள் இன் ஆல்
maram kaḷ in aal
tree pl euph inst

(b) மரத்தினுக்கு (maratinukku) ‘to a tree’

மரம் அத்து இன் கு
maram attu in ku
tree obl euph dat

(c) ஆவினுக்கு (aavinukku) ‘to a cow’

ஆ இன் உ கு
aa in u ku
cow euph euph dat
Nouns in Tamil can also be derived from verbs. However, detailed coverage of deriva-
tional morphology is not included in this work, except the formation of verbal nouns,
which is discussed in Section 2.2.

6
2.1.2 Plurals in Tamil

Tamil noun stems are singular by default. The plural suffix is a bound morpheme in Tamil
that is marked by கள் (kaḷ). However, this marker is also used as an honorary marker in
present-day Tamil usage, especially when added to the third person pronoun அவர்-கள்
(avar-kaḷ) ‘he.3seh’.

2.1.3 Cases in Tamil

Traditional grammarians have identified 8 cases including a vocative (Senavaraiyar, 1938;

Thesikar, 1957). However, modern linguists (Nuhman, 1999; Paramasivam, 2011; Lehmann,
1993) argue that the instrumental case in traditional grammar should be treated as two,
namely instrumental case and sociative case. This modern classification along with the
respective case markers are shown in Table I.6
Some of the case markers in Tamil are free morphemes, specifically the locative, so-
ciative, ablative, and instrumental markers (Lehmann, 1993). For example, சாவியால்
(saaviyaal) ’key.inst’ can also be expressed as சாவி மூலம் (saavi muulam). However,
these free morphemes are not exclusively used for marking cases. For instance, the lexical
meaning of மூலம் (muulam) is ’origin,’ and it is also employed to mark the instrumental
case.
Tamil does not have a definite marker. Instead, definiteness is indicated using demon-
strative markers or by marking irrational objects with the accusative case marker. (Lehmann,
1993). While the object of a sentence may be marked with the accusative case, this is
compulsory only for rational objects (Lehmann, 1993; Nuhman, 1999). Thus, when an
irrational noun carries an accusative marker, it marks that noun as definite, classifying
Tamil as a Differential Object Marking (DOM) language.
6
Thesikar (1957) lists 28 case markers for the locative case and 10 case markers for the vocative case.
Collectively, most of these markers are rarely used in present-day Tamil.

7
 Table I: Case markers in Tamil

Morpheme Morphs / suffixes Example

Nominative - மரம் (maram) ‘a tree’
Accusative ஐ (ai) மரத்ைத (marattai) ‘the tree’
Instrumental ஆல் (aal) மரத்தால் (marattaal) ‘using a tree’
ஒடு, உடன்
Sociative மரத்துடன் (marattuṭan) ‘with a tree’
(oṭu, uṭan)
கு,க்கு,அக்கு,உக்கு
Dative மரத்துக்கு (marattukku) ‘to a tree’
(ku,kku,akku,ukku)
இல்/இன்+இலிருந்து மரத்திலிருந்து (marattiliruntu)
Ablative
(il/in+iliruntu) ‘from a tree’
அது, உைடய, இன்
Genitive மரத்தின் (marattin) ‘of a tree’
( atu, uṭaiya, in)
இல், இடம்
Locative மரத்தில் (marattil) ‘on a tree’
(il, iṭam)
ஆ, ஏ, ஈ
Vocative மரேம (marame) ‘oh a tree!’
(aa,ee,ii)

2.1.4 Nominal paradigm

Rajendran (2009) has proposed a noun paradigm incorporating 26 classes based on their
morphophonological properties. Among these 26 classes, nine classes are used to capture
the morphophonological rules pertaining to pronouns.
Classes of the noun paradigm that are not pronominal are shown in Table II. One word
is selected to represent each class, as Rajendran (2009) does, and the classes are named
on the basis of that representative lexical item. Class distinctions are here determined by
the last vowel modifier (or vowel) [class 1-3,6-8], consonant [classes 4-5,9-16], or whether
it is a two-letter word [class 10 and 11], whether it is a two-letter word and the first letter
ends with a long vowel modifier (or vowel)[classes 13 & 14, 6 & 7]. Based on this, a script
has also been developed by me to classify nouns into their respective classes.7 Classes 6
& 7, and 13 & 14 have been separated into different classes even though they have the
same last character in the orthography. This distinction has been motivated by the fact
that nouns in these classes differ in conjugation patterns. Certain nouns can have several
conjugational forms, which are widely accepted to mark plurality or case. Tamil grammar
texts also accept such multiple forms, and refer it as ேபாலி (pōli) ‘fake’ (Nuhman, 1999;
Thesikar, 1957). For instance, naal ‘day’ shows conjugations of class 4 and class 14 in
present-day Tamil. Therefore, naal can be included naal in both classes.
7
https://github.com/sarves/Tamil-Noun-Classifier

8
2.1.5 Nominal conjugational forms

A total of 36 declension forms are utilised for Tamil nouns, covering plural and case
conjugations, along with external Sandhi markers. Each noun root is acted upon by case
markers in both its singular and plural forms. Furthermore, nouns in their dative or
accusative forms can also be influenced by one of four external Sandhi markers. In sum,
this results in 36 nominal conjugational forms.8

2.2 Verbal Morphology

Tamil verbs have complex morphosyntactic relations, which take auxiliary items such as
question particles, emphatic particles, and conjunctive markers, in addition to regular
inflections such as tense, person, gender, number, and honorifics to build a surface form.
The structure of a simple verb in Tamil is shown in Formula (4).9 The euphonic
particle அன் (an) is optional and when present, it seems to be used to add a dimension
of politeness to the verb in current usage. The medial particle is used to realise tense (past,
present and future), or to negate the verb (Pope, 1979; Lehmann, 1993; Paramasivam,
2011).10 The terminal suffix of a finite verb is used to realise multiple types of information
such as number, person, gender, and rationality (or status) (Pope, 1979; Lehmann, 1993).
However, this terminal suffix cannot be chunked or divided to extract these information
— it is a portmanteau morph. For instance, in (5), -aan denotes that the terminal-suffix
is third person, singular, masculine, and rational. As for other morphosyntactic features,
Tamil has singular and plural values for number, first/second/third person values, and
three gender values: masculine, feminine and neuter. In addition to these three genders,
a fourth class called ‘epicene’ is used to mark the third person plural forms of rational
entities (Lehmann, 1993). This is what affects the choice of the correct terminal suffix.

(4) Verb=lemma-form+<medial-particle>+(euphonic-particle)+<terminal-suffix>

In addition to simple verbs, Tamil also has complex or compound verbs with more than one
verbal root, which may express mood, aspect, negation, interrogative, emphasis, speaker
perspective, conditional, and causal relations (Annamalai et al., 2014). Agesthialingom
(1971) claims that Tamil can have up to four verbal roots in one verb form. For instance,
8
Borrowed words from Sanskrit also can take a nagation marker as the prefix. For instance, நியாயம்
(niyāyam) ‘Justice’ becomes அநியாயம் (aniyāyam) ‘Injustice’.
9
There are two types of transitives in Tamil. The first type is indicated by specific markers that
distinguish between intransitive and transitive forms, which we can refer to as derived transitives. In the
second case, transitivity is inherent to the root itself, and there is no corresponding intransitive form; this
can be can refer to these as inherent transitives.(Agesthialingom, 1971) This work does not distinguish
derived and inherent transitives.
10
Old Tamil has two tenses, past and non-past. There are five allomorphs for past tense: -t-, -nt-,
-in-, -i-, -tt-, and three for non-past tense: -v-, -p -, -pp-.

9
there are four verbal roots in the complex verb in (5): vaa ‘come’, koḷ ‘hold’, iru ‘be’
and iru ‘be’. koḷ ‘hold’ and iru ‘be’ in the middle together signal a continuous aspect.
In (5), we observe that verbal conjugation is only expressed on the last verbal root of
the sequence, where it takes tenses, person, number, and gender (png) marking. The
preceding verbs appear either in a participial or infinitival form.

Table II: The Tamil nominal paradigm

No. Class name Plural Sample case markers

Marker
1 கடா (kaṭaa) ‘male goat’ or பசு க்-கள் ைவ, வால், வுடன், வுக்கு (vai,
(pasu) ‘cow’ (k-kaḷ) vaal, vuṭn, vukku)
2 எலி (eli) ‘rat’ or ெநய் (ney) கள் (kaḷ) ைய, யால், யுடன், யுக்கு (yai,
‘ghee’ yaal, yuṭn, yukku)
3 ஈ (ii) ‘house fly’ க்கள் ைய, யால், யுடன், யுக்கு (yai,
(kkaḷ) yaal, yuṭn, yukku)
4 நாள் (naal) ‘day’ or கால் கள் (kaḷ) ை◌, ◌ா, ◌ு, ே◌ா ( i, aa, u, oo)
(kaal) ‘leg’
5 பலர் (palar) ‘many people’ - ை◌, ◌ா, ◌ு, ே◌ா (i, aa, u, oo)
6 காடு (kaaṭu) ‘forest’ கள்(kaḷ) ட்ைட, ட்டால், ட்டுக்கு, ட்ேடாடு
( ṭai, ṭaal, ṭukku, ṭooṭu)
7 வண்டு (vandu) ‘beetle’ கள் (kaḷ) ை◌, ◌ா, ◌ு, ே◌ா (i, aa, uu, oo)
8 ஆறு (aaru) ‘river’ கள் (kaḷ) ைற, றால், க்கு, ேறாடு (r̲ai, r̲aal,
kku, r̲ooṭu)
9 கண் (kan) ‘eye’ கள் (kaḷ) ைண, ணால், ணுடன், ணுக்கு
(ṇai, ṇaal, ṇuṭan, ṇukku)
10 ெபான் (pon) ‘gold’ கள் (kaḷ) ைன, னால், னுடன், னுக்கு (nai,
naal, nuṭan, nukku)
11 மாணவன் (maanavan) ‘stu- ர்கள் (rḷ) ைன, னால், னுடன், னுக்கு (nai,
dent’ naal, nuṭan, nukku)
12 புல் (pul) ‘grass’ ற்கள் ைல, லால், லுடன், லுக்கு (lai,
(r̲kaḷ) laal, luṭan, lukku)
13 முள் (mul) ‘thorn’ ட்கள் ைள, ளால், ளுடன், ளுக்கு (ḷai,
(ṭkaḷ) ḷaal, ḷuṭan, ḷukku)
14 நாள் (naal) ‘day’ ட்கள் ை◌, ◌ா, ◌ு, ே◌ா (i, aa, uu, oo)
(ṭkaḷ)
15 மரம் (maram) ‘tree’ ங்கள் த்ைத, த்தால், த்துக்கு, த்ேதாடு
(ṅkaḷ) (ttai, ttaal, ttukku, ttooṭu)
16 சுவர் (suvar) ‘wall’ கள் (kaḷ) ற்ைற, ற்றால், ற்றுக்கு, ற்ேறாடு
(r̲r̲ai, r̲r̲aal, r̲r̲ukku, r̲r̲ooṭu)

10
(5) வந்துெகாண்டிருந்திருக்கிறான்
vantukoṇṭiruntirukkiraan
vantu-koṇṭiru-iru-kkir-aan
come.vpart-hold_be.vpart-be-pres-3smr
‘(he) has been coming’

Complex verbs in Tamil can be written as separate tokens, as in (6), or as a single

token, as in (7).

(6) வாங்கச் ெசய்தான்

vang-a-c sei-t-aan
buy-inf-sandhi_c do-past-3smr
‘(he) made someone buy.’
(7) வாங்கச்ெசய்தான்
vang-a-c-sei-t-aan
buy-inf-sandhi_c-do-past-3smr
‘(he) made someone buy.’
(8) வாங்கிக்ெகாடுத்தான்
vāṅkik-koṭu-tt-āṉ
buy-vpart-sandhi_k-give-past-3smr
‘(he) bought for someone’

The set of verbs that form complex verb forms in conjunction with the main verb are
identified and categorised based on their structure and function, primarily relying on the
discussions in Boologarambai (1986). However, further research is necessary to identify
additional complex verbal conjugational forms and their respective functions.

11
 Table III: List of complex verb forms

Verb types Verb roots11 Structure of a finite verb

Aspectual இரு (iru) ‘be’, main-verb+vp
verbs ெகாண்டிரு (koṇṭiru) ‘keep’, +aspectual-verb+tense+png
விடு (viṭu) ‘leave’
முடி (muṭi) ‘finish’
Attitude verbs ேபா (po) ‘go’ main-verb+vp
ேபாடு (poṭu) ‘put’, +attitude-verb+tense+png
தள்ளு (taḷḷu) ‘push’,
தீர் (tiir) ‘solve’,
ெதாைல (tolai) ‘get lost’
Non-attitude இடு (iṭu) ‘put’, main-verb+inf +non-attitude-
or Light Verbs ெகாடு (koṭu) ‘give’, verb+tense+png
பார் (paar) ‘see’,
வா (vaa) ‘come’
Modal verbs ேவண்டு (veeṇṭu) ‘want’, main-verb+inf
கூடு (kooṭu) ‘may’ +modal-verb+tense+png
Causatives பண்ணு (paṇṇu) ‘make’, main-verb+inf
ெசய் (sei) ‘make’, +causative-verb+tense+png
ைவ (vai) ‘cause’
Passivisers படு (paṭu) ‘suffer’, main-verb+inf
ெபறு (peru) ‘get’ +passiviser+tense+png

2.2.1 Verbal paradigm

Tamil verbs can be classified on the basis of criteria that can be either morphological,
syntactic or semantic (Paramasivam, 2011). Many scholars, including Lisker (1951), Graul
(1855), and Arden (1910) have classified verbs on the basis of what their morphophonemic
changes they display as part of their conjugations. Graul (1855) has provided an early
classification on which other scholars have built their proposals, including Irākavaiyaṅkār
(1958) and Sithiraputhiran (2004). Graul’s classification has also been adopted for the
Tamil lexicon project (Rajaram, 1986). This classification of Tamil verbal lemmas includes
12 categories or classes and is based on the tense markers as displayed on the verbs. As
shown in Table IV, when verb forms are conjugated for tense markers, these are not
just added to the lemma form via concatenation. Rather changes or new letters may be
introduced, and these constructions are handled with the use of alternation rules. For
4
Note: I have given the literal meaning of the verbs in the column - Verb roots. However, when
functioning as an auxiliary or light verb, these meanings may not hold.

12
instance, in class 12, when the past tense marker is coined, it is not just added to the
lemma, e.g. (நட+த் (naṭa+t) ‘walk.sandhi_t. Instead, the letter ந் (nt) is inserted as
in நட+ந்+த் (naṭa+n+t). A similar behaviour presents in other classes as well.
In addition to these classes, I have identified five irregular verbs that do not fit into
the paradigm given in Table IV: காண் (kaaṇ) ‘see’, வா (vaa) ‘come’, சா (saa) ‘die’, தா
(taa) ‘give’, ேவ (vee) ‘boil’.

Table IV: The Tamil Verbal

No. Class name Past Present tense marker Future tense

tense marker
marker
1 ெசய் (sei) த் (t) கிற், கின்ற் (kir, kinr) வ், உம் (v, um)
2 ஆள் (aaḷ) ட் (ṭ) கிற், கின்ற் (kir, kinr) வ், உம் (v, um)
3 ெகால் (kol) ற் ( ṛ) கிற், கின்ற் (kir, kinr) வ், உம் (v, um)
4 கடி (kaṭi) த் (t) கிற், கின்ற் (kir, kinr) வ், உம் (v, um)
5 அஞ்சு (angu) இன் (in) கிற், கின்ற் (kir, kinr) வ், உம் (v, um)
6.1 அடு (aṭu) ட் (ṭ) கிற், கின்ற் (kir, kinr) வ், உம் (v, um)
6.2 நகு (naku) க் (k) கிற், கின்ற் (kir, kinr) வ், உம் (v, um)
6.3 உறு (uru) ற் (ṛ) கிற், கின்ற் (kir, kinr) வ், உம் (v, um)
7 உண் (uṇ) ட் (ṭ) கிற், கின்ற் (kir, kinr) ப், உம் (p, um)
8 தின் (tin) ற் ( ṛ) கிற், கின்ற் (kir,kinr) ப், உம் (p, um)
9 ெகாள் (koḷ) ட் ( ṭ) கிற், கின்ற் (kir,kinr) ப், உம் (p,um)
10 நில் (nil) ற் ( ṛ) கிற், கின்ற் (kir,kinr) ப், உம் (p, um)
11 அபகரி (abakari) த் (t) க்கிற், க்கின்ற் (kkir, ப்ப், உம் (pp, um)
kkinr)
12 நட (naṭa) த் (t) க்கிற், க்கின்ற் (kkir, ப்ப், உம் (pp, um)
kkinr)

2.2.2 Verbal conjugational forms

Annamalai et al. (2014) have identified 254 forms for each Tamil verb following a rigorous
analysis of their corpus of contemporary texts. Some verbs may, however, not take all of
the 254 forms. Rajaram (1986) has identified 21 forms for each verb from a pedagogical
perspective. On the other hand, Kumar et al. (2010) claims that a Tamil verb lemma
can take up to 8,000 forms if derivations are also considered. Sarveswaran et al. (2021)
has compiled over 3,000 verb lemmas and identified close to 600 forms for each verb
lemma. For example, 3.13 shows 582 forms for the verb lemma நட naṭa ’walk’, including
the suffixation of Sandhi. This count excludes the causative derivation lemma நடத்து
naṭattu ’make walk/conduct’. Additionally, this list does not account for the suffixation

13
of auxiliary verbs, including modal or aspect aspects. When considering auxiliary verbs
as well, each verb can potentially yield several thousand forms. Sarveswaran et al. (2021)
has made efforts to create these compounds, compiling over 57 million of them for all
3,000+ verb lemmas.12 While some of these words may not be in current use or not used
such formations remain possible.

2.3 Adjectival Morphology

In Tamil, everything which precedes and modify a noun is called ெபயரைட peyaraṭai.
This includes adjectives, nominal modifiers, demonstrative pronouns, numeric modifiers,
and adjectival participles. There are two sorts of adjectives in Tamil: pure adjectives
and derived adjectives. Pure adjectives are words that are used as adjectives without
requiring any suffixation. The attributive type of adjectival derivation, which is formed
by adding the suffix ஆன -aana ‘an adjectiviser’ to nouns, as shown in (9). This suffix is
derived from a verb ஆகு -aaku ‘become’. உள்ள -uḷḷa ‘in’ (rough translation) is the other
adjectiviser used in Tamil (Nuhman, 1999).

(9) உயரமான ���

uyaram-aana malai
tall-be mountain
‘tall mountain’

Example in (10) shows an instance of adjectival modification, where the verb come.past
take an adjectival suffix a to modify the noun. This follows the structure of relative clause.
However, instead of a relative pronoun, here we have a noun. Further, in relative clause,
we mark a as a relativiser.
Example in (11) shows a nominal modifier. Similarly, we can also have numeric mod-
ifiers precedes to nouns.

வந்த ைபயன்
vaa.nt.a paiyaṉ
(10)
come.past.adj boy
‘The boy who came’
அைமதிப் பைட
amaiti-p paṭai
(11)
peace.sandhi_p force
‘A peace force’

Tamil has three demonstratives to indicate spatial deixis: இந்த inta ‘this’, அந்த
anta ‘that’, and உந்த unta ‘between this and that’, which are marked by the respective
demonstrative markers in Tamil: இ i, அ a, and உ u, respectively. Among these markers,
12
https://www.kaggle.com/datasets/sarves/tamilverbs

14
unta is primarily used in the northern region of Sri Lanka — Jaffna, and its usage has
diminished in other parts of the world.13
Adjectives can be reduplicated to modify plural nouns as shown in Example (12),
which is taken from (Nuhman, 1999). Further, superlative forms are derived by addition
modifiers to adjectives in Tamil, as shown in (13). However, in poetic writing this can
even be preceded by மிக mika ‘very’ to mark much more tininess.

சிறிய சிறிய வீடுகள்

ciṟiya ciṟiya vīṭukaḷ
(12)
small small houses
‘small small houses’
சின்னஞ்சிறிய வீடு
ciṉṉañ-ciṟiya vīṭu
(13)
small-small house
‘The smallest house’

2.4 Morphology of adverbs

Everything which precedes and modify a verb is called விைனயைட viṉaiyaṭai, in Tamil.
Similar to adjectives, adverbs in Tamil come in two types: pure adverbs and derived
adverbs. Pure adverbs are word forms that are themselves used as adverbs without
requiring any sort of suffixation; for instance, in (14), the adverb nēṟṟu ‘yesterday’ does
not have any suffixation on it. Several adverbs are subsumed under this type, namely,
temporal, frequency, place, degree, and affirmation (Nuhman, 1999). Manner adverbs fall
under the second type, i.e. derived adverbs. These derived adverbs are formed by adding
an adverbial suffix ஆக/ஆய் -āka/āy ’become’ to nouns, as shown in (15).

ேநற்று வந்தான்
nēṟṟu vantāṉ
(14)
yesterday came (he)
‘He came yesterday’
ேவகமாக வந்தான்
veham-āka vantāṉ
(15)
speed-ADV came (he)
‘(He) came fast’
In this section, the morphology of the Tamil language has been discussed, particularly
focusing on the inflectional morphology of nouns, verbs, adjectives, and adverbs. Brief
mention has also been made of derivational morphology where relevant. It is now evident
13
Ronald E. Asher, in the preface of Suseendirarajah (1999) claims that knowledge of Tamil, specifically
Jaffna Tamil, a variety spoken in Sri Lanka, is a vital element that forms part of the understanding of
the Dravidian family of languages. This is particularly so since certain features of the language have been
preserved in Jaffna Tamil that have been lost elsewhere.

15
that Tamil exhibits a complex morphological system.

3 Syntax of Tamil
Over the years, Tamil has been in contact with several other languages. It is no longer
spoken in some of the countries in which it was spoken several decades ago. Further, due
to the migration that has been happening, Tamil is now spoken all over the world and
has been in contact with a number of Western languages as well. Apart from this array of
language contact, the grammaticalisation of various words in Tamil through time makes
the study more complex. There are, however, only a small number of works that relate to
modern Tamil grammar. Similarly, there is no comprehensive grammar for modern Tamil
written after the 13th century naṉṉūl grammar. It is the derived grammar that is based
on naṉṉūl (Thesikar, 1957) that is used as a high school Tamil textbook (Nuhman, 1999).
Tamil requires a significant amount of proper linguistic study to understand and describe
its grammar, especially syntax of modern Tamil. There has been an attempt to develop
a modern grammar for Tamil by several scholars in Tamil Nadu, including E. Annamalai.
However, no parts of the grammar have been released yet.
This section highlights certain syntactic structures of Tamil extracted from a corpus.
Furthermore, various syntactic analyses in grammar books and research papers have been
followed, although they also require further linguistic exploration. The use of modern
linguistic theories has shed some light on a few constructions such as complex predicates
and light verbs, as detailed in Sarveswaran and Butt (2019), as a better understanding of
these is essential for a complete understanding of their structure and semantics.
Tamil is a highly free word order language, although Subject-Object-Verb (SOV) is the
commonly used constituent order. For this reason, and for further simplification, variants
other than SOV are not covered in this overview.

3.1 Nouns and their structures

Nouns in Tamil are marked for number and case. The number values available are singular
and plural. Case14 in Tamil has received much attention but has evolved over time.
Modern Tamil grammarians propose nine cases; some are discussed in detail below. There
14
The analysis of whether the given case marker is a ‘true’ case marker or ‘just’ a postposition has
not been conducted. The approach has been to follow the existing literature, although there are in-
stances where further analysis may be needed. Nonetheless, this issue has been a longstanding concern
in the Tamil case system (Schiffman, 2004; Caldwell, 1998). In fact, Caldwell (1998) notes the following
regarding the Dravidian case system: “All case-relations are expressed by means of postpositions, or
postpositional suffixes. In reality, most of the postpositions are separate words; in all the Dravidian
dialects, they retain traces of their original character as auxiliary nouns. Several case signs, especially in
the more cultivated dialects, have lost the faculty of separate existence and can only be treated now as
case terminations. However, there is no reason to doubt that they are all postpositional nouns originally.”

16
is evidence that some case forms in Tamil are also used to mark other cases. For instance,
the accusative case can sometimes be used to express the instrumental (Joseph, 1893).
This section will not make any further reference to this additional syncretic complexity.
These aspects require further linguistic exploration.
Apart from numbers and case, functional elements like adverbials, postpositions, par-
ticles, and clitics can also be attached to nouns. Adverbials are primarily of a spatial or
temporal nature. While clitics can be added to nouns irrespective of their number and
case marking, particles can be attached to nouns if they are in the nominative, accusative
or dative case. Adverbials and postpositions are attached to nouns if they are in the
nominative or the dative case.

3.1.1 Nominative case

Syntactically unmarked, bare nouns are understood to be nominative-marked. These may

function as a subject, predicate, subject-complement, objective-complement, or object
in Tamil. Except for dative subjects, subjects in nominative cases usually necessitate
subject-verb agreement, where the verb shows agreement with the nominative case marked
nominal in person, number, gender (and rationality).

கண்ணன் ஒரு மாணவன்

kaṇṇaṉ oru māṇavaṉ
(16)
Kannan.nom a student.nom.masc
‘Kannan is a student.’
குமார் தைலவன் ஆனான்
kumār talaivaṉ āṉāṉ
(17)
Kumar.nom leader.nom.masc become.past.3smr
‘Kumar became a leader.’

Example (16) is a nominal predicative construction — an equative construction. Such

constructions are commonly used in Tamil, where the linking verb is dropped in the present
tense constructions. In such null-copula constructions, order of phrases play a role and
determine the subject. However, for other tenses, a copula āna ‘become’ is introduced
to carry tenses, as in (17). The other variant of this is shown in (18) where āka — an
adverbial suffix — is added to the nominal predicate in (17). Although examples (17) and
(18) following are structurally different, quantifying the difference has proven challenging;
it necessitates linguistic exploration. In (18), ’āka’ is glossed as ’adv’ because it serves as
an adverbial suffix used to convert nouns into adverbs.

குமார் தைலவனாக ஆனான்

kumār talaivaṉāka āṉāṉ
(18)
Kumar.nom leader.masc.adv become.past.3smr
‘Kumar became the leader.’

17
 குமார் இராமைனத் தைலவன் ஆக்கினான்
kumār irāmaṉait talaivaṉ ākkiṉāṉ
(19)
Kumar.nom Raman.acc.sandhi-t leader.nom make.past.3smr
‘Kumar made Raman a leader.’
In N+V predication – light verb construction – N is always in the nominative case
and serve as the semantic head of the phrase. For instance, (18) is a resultative structure
with an N+V predicate, where N is in the nominative case.

3.1.2 Accusative case

The case marker -ai is used to mark the accusative case in Tamil. Like other case markers,
this is also added to the nominal stem, oblique stem,15 or a pluralised noun.
Tamil has differential object marking, where objects are marked with and without an
accusative case on the basis of rationality. It is obligatory to have the accusative case
marking for rational entities as shown in (19). In his work, (Lehmann, 1993) notes that
there are a few exceptional instances where rational entities do not take the accusative
case marker when functioning as objects. However, no such instances were found in our
corpora.
The accusative marking is optional for irrational entities as shown in (20), and adding
an accusative case marker onto these irrational entities ends up denoting definitiveness
as in (21). However, if the accusative case is not overtly marked, as in (20), the order of
the object becomes restricted in relation to other constituents. Basically, the object has
to follow the subject; otherwise, the sentence becomes ambiguous.

குமார் பந்து அடித்தான்

kumār pantu aṭittāṉ
(20)
Kumar.nom ball.nom hit.past.3smr
‘Kumar hit a ball.’
குமார் பந்ைத அடித்தான்
kumār pantai aṭittāṉ
(21)
Kumar.nom ball.acc hit.past.3smr
‘Kumar hit the ball.’

3.1.3 Dative case

The dative case marked by உக்கு -ukku has many different functions in Tamil; Lehmann
(1993) lists nine different functions. However, the following four different instances of
dative marking were found in the corpus:
1. Indirect objects
15
Some nouns take a suffix called oblique suffix, which allows the root to take other suffixes.

18
 Ditransitive verbs in Tamil take indirect objects marked with dative. Unlike the
accusative case, the case suffix is always added irrespective of rationality.

இது அரசுக்கு ஒரு சவால் அல்ல

itu aracukku oru cavāl alla
(22)
this state.dat a challenge.nom not
‘This is not a challenge to the state.’
2. Benefactive case
On top of -ukku, a suffix -āka is added to mark the benefactive case in Tamil. Although
this can be considered a separate construction type, since this is somewhat related to the
dative marking this is covered under the dative.

நாம் நாட்டுக்காக ேசைவ ெசய்யும் குழு

nām nāṭṭukkāka cēvai ceyyum kuḻu
(23)
we.nom country.ben service.nom do.fut.rel group.nom
‘We are a group that serves the country.’
3. Dative subjects
While some scholars consider dative subjects as indirect objects, others, such as Mo-
hanan and Mohanan (1990), and Pappuswamy (2005), accept the concept of dative sub-
jects. This phenomenon is found in other languages as well (Butt et al., 2006). Unlike
nominative subjects, dative subjects show mixed behaviours. They, for instance, do not
show all the subject properties shown by a nominative subject, including the person-
number-gender agreement with the verb.
Verbal predicates with dative subjects marked for third-person, neuter-gender, and
singular — are also referred to as default agreement — irrespective of subject’s number,
gender, rationality, and person. This default agreement is marked using உம் -um, as
shown in (24). In this construction, the object does not carry the accusative marking.
However, from other constructions it is evident that objects can carry overt accusative
marking.

எனக்குச் சிங்களம் ெதரியும்

eṉakkuc ciṅkaḷam teriy-um
(24)
I.dat.sandhi_c Sinhala.nom know.fut-3sn
‘I know Sinhala.’
Nominal predicates, which represent feelings, sensations, or states of being, with dative
subjects also widely used in Tamil. For instance, (25) shows a nominal predicate example
with a dative subject.

19
 எனக்குப் பசி
eṉakku-p paci
(25)
I.dat-sandhi_p hungry.nom
‘I am hungry’
4. Goal marker
Case marker -ukku is also used to mark the goal argument of motion verbs in Tamil,
as shown in (26). This is analysed as an oblique argument in grammatical framework like
the Universal Dependencies and Lexical Functional Grammar.

அண்ணா கண்டிக்குச் ெசன்றான்

aṇṇā kaṇṭikkuc ceṉṟāṉ
(26)
elder-brother.nom kandy.dat.sandhi_c go.past.3smr
‘Elder brother went to Kandy.’

3.1.4 Instrumental case

The instrumental case is marked by -āl to show instrumentation as in (27). In addition,

the instrumentation can also be marked via the use of the postposition மூலம் mūlam
‘with which’, which may or may not be suffixed to the root. In other context, mūlam also
means ‘a source’.
Apart from this regular use-case, the instrumental case is also used to mark the agent
in passive constructions, as shown in (28), and to form conditional verbal clauses, as
discussed later in this chapter.

ைபயன் சாவியால் கதைவத் திறந்தான்

paiyaṉ cāviyāl katavait tiṟantāṉ
(27)
boy.nom key.ins door.acc.sandhi_t open.past.3smr
‘A boy opened the door with a key.’
அைமச்சரைவயால் புதிய தீர்மானங்கள் முன்ெனடுக்கப்பட்டன
amaiccaravaiyāl putiya tīrmāṉaṅkaḷ muṉṉeṭukkappaṭṭaṉa
(28)
cabinet.ins new resolutions.nom forward-take.pass.past.3pln
‘New resolutions were taken-forward by the Cabinet.’

3.1.5 Locative case

In modern Tamil, the -il marker is used to mark the location in space and time, and mode
on irrational and rational nouns (Lehmann, 1993). For instance, (29) shows how -il is
used to mark the location in space. -iṭam is another marker used to mark locative cases
only on rational nouns. Lehmann (1993) claims that -iṭam expresses the goal of motion,
transaction source, emotion target, and temporary possession. For instance, my corpora
consist (30), which is an example of a goal of motion of speech.

20
 குமார் கூட்டத்தில் ேபசினான்
kumār kūṭṭattil pēciṉāṉ
(29)
Kumar.nom meeting.loc speak.past.3smr
‘Kumar spoke at the meeting.’
குமார் கண்ணனிடம் ேபசினான்
kumār kaṇṇaṉiṭam pēciṉāṉ
(30)
Kumar.nom Kannan.loc speak.past.3smr
‘Kumar spoke to Kannan.’

3.1.6 Ablative case

-iruntu is used to mark ablative cases on top of locative-cased nouns. For instance, -il
+ -iruntu = -iliruntu would function as ablative case marker for irrational nouns, as in
(31). Similarly, -iṭam + -iruntu = -iṭamiruntu in turn functions as the marker for rational
nouns.

நான் ெகாழும்பிலிருந்து வீட்டிற்கு வந்ேதன்

nāṉ koḻump-il-iruntu vīṭṭiṟku vantēṉ
(31)
I.nom Colombo.loc.abl home.dat come.past.1sr
‘I came home from Colombo.’

3.1.7 Genitive case

உைடய -uṭaiya, அது -atu, இன் -iṉ are the markers used to mark genitive case in Tamil,
which is available in the context of possession, and to show a thing’s source or a charac-
teristic/trait of something. There is no exact rule to say which marker needs to be used
in what context (Schiffman, 2004). (32) shows how -iṉ is used to mark possessive. -atu
is a confusing maker, which is also used to mark third-person, singular, and neuter in
verbs and serves as a pronoun to mark the same. In addition, -atu functions differently in
relative clause constructions. A genitive case example using -atu is shown in (33), as in
this example, words can be shuffled when -atu is used to mark the genitive case. However,
it is not possible with other genitive markers.

நான் அவரின் பிள்ைள

nāṉ avariṉ piḷḷai
(32)
I.nom he-hon.gen child.nom
‘I am his child.’
பள்ளிக்கூடம் குமாரது
paḷḷikkūṭam kumāratu
(33)
school.nom Kumar.gen
‘Kumar’s school.’
Further, it is very common in Tamil that these genitive markers are dropped if it does
not lead to any ambiguity. For instance, see (34), which is comparable to (32), where

21
genitive case marker -iṉ is dropped, yet it shows the possessiveness.

நான் அவர் பிள்ைள

nāṉ avar piḷḷai
(34)
I.nom he-hon.nom child.nom
‘I am his child.’

3.1.8 Sociative case

ஒடு -oṭu and உடன் -uṭaṉ are the markers used to mark sociative cases in Tamil, see (35).
When multiple entities function together as a subject, this conjunction will be reflected
in subject-verb agreement, which is discussed later under topic of Coordination (Section
3.7).

நான் அப்பாவுடன் ேபாேனன்

nāṉ appāvuṭaṉ pōṉēṉ
(35)
I father.soc go.past.1s
‘I went with father.’

3.2 Agreement
A nominal subject and verbal predicate agree for rationality, gender, number, and person
in Tamil. Tamil grammar textbooks, e.g. Nuhman (1999), claim that there are five gender
values, as given below. Although this classification is referred to as gender, the markers
do not express gender values alone. Rather it is gender + number + rationality values
that are expressed.

1. āṇpāl - masculine singular

2. peṇpāl - feminine singular
3. palarpāl - rational plural
4. oṉṟaṉpāl - irrational singular
5. palaviṉpāl - irrational plural

Four gender values can also be employed - masculine, feminine, epicene,16 and neuter,
without delving into the added complexity of combining gender and number markings
being fused onto the same form.
Plurality agreement, although it is obligatory, not found in some of the instances in
the corpus for irrational nouns. It is unclear whether this discrepancy may be related
to regional differences. For example, (36) was found in the corpus — this may be from
16
Epicene is used to denote singular-rational entities that are not classified as either masculine or
feminine. (Lehmann, 1998)

22
Indian Tamil, although the alternative structure otherwise available in the Sri Lankan
context is the one in (37).

மூன்று நாய்கள் வந்தது

mūṉṟu nāykaḷ vantatu
(36)
three dog.pl come.past.3sn
‘Three dogs came.’
மூன்று நாய்கள் வந்தன
mūṉṟu nāykaḷ vantaṉa
(37)
three dog.pl come.past.3pln
‘Three dogs came.’
When it comes to person, there are three values, namely, first, second, and third,
as in most other languages. However, Tamil has one more value that is used to mark a
deictic reference between the second and the third person. Although special pronouns
are used to address this medial deictic, this does not influence any syntactic functions, at
least in the corpora used.
In addition to the three main person values, Lehmann (1993) has proposed the fourth
person. According to him, this fourth person pronoun17 — tāṉ — is always co-referential
with the subject of the same clause or higher. In addition, there is no special agreement
displayed between this fourth-person pronoun and the verb. On the other hand, this -tāṉ
can easily be confused with the -tāṉ clitic that is an emphatic marker in Tamil. Based on
the data, tāṉ is treated as a clitic when it is written together with the headword, as in
(38). Otherwise, when tāṉ is written as a separate word, it functions as a fourth person
pronoun.

நான் ெசய்தது தவறுதான்

nāṉ ceytatu tavaṟutāṉ
(38)
I.nom do.past.rel wrong.nom.emph
‘What I did was wrong.’
கண்ணன் தான் பரீட்ைசயில் சித்தியைடயமாட்டான்
(39) kaṇṇaṉ tāṉ parīṭcaiyil cittiyaṭaiya.māṭṭ.āṉ
Kannan.nom himself exam.loc pass.will-not.3smr

என்று நிைனத்தான்
eṉṟu niṉaittāṉ
that think.past.3smr
‘Kannan just thought he would not pass the exam.’
In Tamil, although there is a specific suffix with which to mark honorifics on ver-
bal predicates, it is common to use the plural of suffix as a means to mark honorifics.
Therefore, agreement for number may not always hold between the subject and the verbal
17
Although I believe that tāṉ is a reflexive pronoun, I have used the terminology of Lehmann (1993)

23
predicate when it comes to honorifics, as shown in (41).

நான் உன்ைனப் ேபாகவிடமாட்ேடன்

nāṉ uṉṉaip pōka.viṭa.māṭṭ.ēṉ
(40)
I you.acc.sandhi_p go.let.will-not.1s
‘I will not let you go.’
ஜனாதிபதி கூட்டத்தில் ேபசினார்கள்
jaṉātipati kūṭṭattil pēciṉārkaḷ
(41)
president.nom meeting.loc speak.past.3pler
‘The President spoke at the meeting.’

3.3 Negation
Negation can be realised morphologically using the -ā or māṭṭu morphs, or by syntactically
using the word illai. -ā is used to negative all types of verbal constructions, including
finite and conditionals, as shown in (42). -aa can easily be confused with the question
particle. mattu is specified as a future negative marker, as shown in (43) (Nuhman, 1999;
Lehmann, 1993).

நீ வராமலிருந்தால் அப்பா ேகாபிப்பார்

nī varāmaliruntāl appā kōpippār
(42)
you come-not-if father.nom get-angry.fut.3ser
‘Dad will get angry if you don’t come.’
நான் உன்ைனப் ேபாகவிடமாட்ேடன்
nāṉ uṉṉaip pōka.viṭa.māṭṭ.ēṉ
(43)
I you.acc.sandhi_p go.let.will-not.1s
‘I will not let you go.’
The word illai can function as a habitual (44), existential (45), and possessive (46)
negator — as shown in the examples.

நான் அதிகாைலயில் எழும்புவது இல்ைல

nāṉ atikālaiyil eḻumpuvatu illai
(44)
I morning.loc get-up no
‘I do not wake up early.’
அவனுக்குப் புத்தி இல்ைல
avaṉukkup putti illai
(45)
he.dat.sandhi_p wit.nom no
‘He has no wit.’
எனக்கு சந்ேதகம் இல்ைல
eṉakku cantēkam illai
(46)
I.dat doubt.nom no
‘I have no doubt.’

24
3.4 Postpositions in Tamil
Postpositions are sometimes referred to as particles (Arden, 1910), and are the equivalent
to prepositions in English (Schiffman, 2004), except for their varied syntactic position,
since postpositions are suffixed onto the words which they govern. However, analysing
and listing all postpositions in Tamil is difficult since most of the postpositions are often
nouns or verbs in their origin. Almost any verb in the language can be advanced to
candidacy as a postposition (Schiffman, 2004). For instance, pōṭu ‘put’ and koḷ ‘hold’
can be used as postpositions with which to mark instrumental case (Schiffman, 2004) in
modern usage. For example, (47) shows how koḷ ‘hold’ is used as a postposition to mark
the instrumental case. However, if one can look deeper, it is clear that koḷ still functions
as a verb and give the object case to stick. Otherwise, there are no ways to explain the
case marking on the stick. Therefore, postpositions in Tamil have not been discussed in
this version as they require further linguistic exploration. This will be explored in future
revisions, step by step.

ெபாலீஸ் தடிையக்ெகாண்டு கூட்டத்ைத விரட்டியது

polīs taṭiyaik-koṇṭu kūṭṭattai viraṭṭiyatu
(47)
police stick.acc-using crowd.acc chase.past.3sn
‘Police chased the crowd with stick.’

3.5 Verbs and their syntactic structures

Tamil verbs have been mostly analysed from a prescriptive perspective, and most of
these studies are based on the very first Tamil grammar called tolkāppiyam and a derived
piece of work, the naṉṉūl (Thesikar, 1957), published in the 13th century CE. From the
18th century CE onwards, Western scholars have also contributed to the study of Tamil
grammar. However, except for the attempt by Annamalai (2013), none of the scholars has
clearly articulated the differences between complex predicates, serial verb constructions
(Steever, 2005; Fedson, 1981), complex verbs (Agesthialingom, 1971), and compound verbs
(Agesthialingom, 1971; Nuhman, 1999; Fedson, 1981; Paramasivam, 2011).

3.5.1 Complex Predicates

The study of complex predicates (CP) has received a great deal of attention in the lin-
guistic literature and a number of distinct interpretations. This research is based on
the definition proposed in Butt (1995), which views CPs as being formed when two or
more predicational units enter into a relationship of co-predication. Each predicational
unit adds arguments to a mono-clausal prediction; a similar definition or idea can also
be found in Mohanan (1994) and Alsina et al. (1997). It is important to identify CPs

25
to understand the differences with respect to other potentially confusing categories for
the development of computational resources such as computational grammars (Butt and
King, 2002), WordNet (Chakrabarti et al., 2007), and machine translation (Kaplan and
Wedekind, 1993; Butt, 1994).
Complex predicates are very common in Tamil (Annamalai, 2013). For instance, verbs
like ைவ (vay) ‘place’, விடு (vidu) ‘let go’, பார் (paar) ‘see/look’ may function both as
main/full and light verbs. As light verbs, they mean ‘cause’, ‘let’ and ‘try’, respectively
(Annamalai, 2013).
As noted in the existing literature (Annamalai, 2013; Steever, 2005; Lehmann, 1993;
Rajendran, 2004), Tamil is well known for a diverse type of Verb-Verb (V-V) and Noun-
Verb (N-V) constructions.

• verb+verb constructions involve a series of auxiliary or light verbs that are added
periphrastically to the first verb, either in the form of a verbal participle or an in-
finitive. Muthuchchanmugan (2005) shows that up to four verbal units can follow
the main verb (also referred to as the lexical headword) in Tamil. However, as he
claims, whether all of these are auxiliaries is debatable. In V-V constructions, the
terminal verbal unit is the final item in a sequence. The preceding verbal units can
be in either an adverbial or infinitival form. The terminal verbal unit is the item
that carries all the functional information, such as tense, person, number, and gen-
der. The V-V sequences are used to express a range of semantic information. This
includes cross-linguistically well-established categories such as the causative, pas-
sive, permissive, negation, aspectual information, and mood and modality, includ-
ing obligation vs. possibility. The literature also describes definitive and conclusive
meanings, the expression of irritation, carelessness, augmentation, prediction and
intention (Paramasivam, 2011; Muthuchchanmugan, 2005). For instance, Example
(48) shows construction with definite conclusive meaning, where the muṭi ‘finish’ is
the head verb that gives the meaning of complete and the auxiliary verb viṭu ‘leave’
(in the example, viṭu has become viṭ due to the phonological transformation) marks
the definite completion.

• noun+verb constructions where the noun function as the head and the verb as a
light verb. (49) is an example for such a construction where kaitu ‘arrest’ is the
nominal head and cey ‘do’ is a light verb.

அைமச்சர் கூட்டத்ைத முடித்துவிட்டார்

amaiccar kūṭṭattai muṭittu-viṭṭār
(48)
minister.nom meeting.acc conclude-definite.past.3ser
‘The Minister concluded the meeting.’

26
 ெபாலீஸ் கள்வைனக் ைகது ெசய்தது
polīs kaḷvaṉaik kaitu ceytatu
(49)
police thief.acc.sandhi_k arrest.nom do.past.3sn
‘The police arrested the thief.’

3.5.2 Light Verb Construction

Light Verbs (LV) are also complex predicates. They differ from main/full verbs in their
syntactic distribution and lexical semantics. While main verbs can stand alone and pred-
icate independently, light verbs depend on the existence of another predicative element
in the clause. LVs are light in the sense that they do not carry the meaning of the corre-
sponding full verb, yet they still contain lexical-semantic information (Butt, 2010). Unlike
auxiliaries, they are not fully functional elements. The light verb forms a syntactically
monocausal unit with the main predicational element. Following Butt (2010), we can
analyse LVs as a separate syntactic category and differentiate them from both main verbs
and auxiliaries in the language. Such structures are common in South Asian Languages
(Butt and Lahiri, 2013), including Tamil (Annamalai, 2013).
Annamalai (2013) has analysed various V-V, Infinitive-V, N-V and Verbal Participle-
V sequences that have been analysed as Serial Verb Constructions (SVC) and Complex
Predicates (CP).
Example (50) illustrates a simple transitive verb வாங்கு (vangu) ‘buy’. An example
of an N+V structure is given in (49).
The same main verb used together with ெகாடு (kodu) ‘give’ in its light verb sense forms
a CP in (51). The light verb ‘give’ contributes a beneficiary meaning to the predication
and licenses the use of an additional beneficiary indirect object (obj-th). The light verb
as the terminal verbal unit carries functional information, which in this case has to do
with tense, number and person.

நான் காைர வாங்கிேனன்

naan carai vanginen
(50)
I.nom car.acc buy.past.1s
‘I bought the car.’

நான் அவனுக்குக் காைர வாங்கிக்ெகாடுத்ேதன்

naan avanukku-k carai vangikkoduththen
(51)
I.nom he.dat-san car.acc buy.vp.san.give.past.1s
‘I bought him a car.’

Example in (52) shows an alternative version of (51), in which the two parts of the
complex predication are realised together. The Sandhi [k] is also triggered on the main
verb வாங்கு (vangu) ‘buy’ just as in the single word realisation in (51), thus further
consolidating the monocausal function of the whole construct.

27
 நான் அவனுக்குக் காைர வாங்கிக் ெகாடுத்ேதன்
naan avanukku-k kar-ai vangi-k koduththen
(52)
I.nom he.nom-san car-acc buy.vp.san give.past.1s
‘I bought car for him.’

3.5.3 Causatives

Causative verbs indicate that one person or thing causes another to do something for an-
other person. A causative in Tamil can be realised either morphologically or syntactically.

1. The morphological realisation of causation in Tamil

The causation in Tamil can be morphologically realised via three morphs: வி vi, பி
pi, and ப்பி ppi occur before the tense maker in a verb (Steever, 2005). For instance,
(53) shows how the causative marker வி vi is used to causative a verb. The choice of the
causative morph depends on the last vowel of the verbal root and is thus phonologically
conditioned.

வாங்குவித்தான் (vanguvittaan)
வாங்கு -வி -த் -த் -ஆன்
(53) vangu -vi -t -t -aan
buy -caus -san -past -3smr
‘he made somebody buy (something)’

2. The syntactical realisation of causation in Tamil

The syntactical realisation of causation in Tamil can also be realised by adding one
of the following verbs after the infinitive form of the main verb: ெசய் sei ‘do’, ைவ vai
‘put’, பண்ணு pannu ‘do’. In this case, these verbs do not predicate as full verbs and have
the character of light verbs, expressing the non-referential meaning ‘make’, as shown in
(54).

அவைன ஒரு கார் வாங்கச் ெசய்ேதன்

avanai oru car vanga-c seithen
(54)
he.acc a car.nom buy.inf.san make.past.3sm
‘(I) made him buy a car.’
Double causatives also exist in Tamil, where the causative form of a derivational verb
can take an additional causative morphological marker to indicate double causation. This
will be explored further in the next version of this research communication.

3.5.4 Passives in Tamil

Passive constructions in Tamil are realised via a V-V construction, and the verb படு paṭu
‘be touched/be experienced/sleep’ is used to passivise constructions, where padu functions

28
as an auxiliary verb.
Together with an infinitive form of the main verb, it gives the meaning of ‘be subjected
to’. For instance, (55) is a passive construction where paṭu is added to the infinitival form
of vāṅku to passivise the construction. Further, agent ‘rām’ become an instrumental
oblique and object ‘car’ has become nominative subject.

ராமால் கார் வாங்கப்பட்டது

rāmāl kār vāṅkappaṭṭatu
(55)
ram.inst car.nom buy.inf.pass.past.3sn
‘A car was bought by Ram.’
ராம் அவனுக்கு ஒரு கார் வாங்கிக் ெகாடுத்தான்
ram avanukku oru car vangki-k koduththaan
(56)
ram.nom he.dat a car.nom buy.vp.san give.vp.past.3sm
‘Ram bought a car for him.’

ராமால் அவனுக்கு ஒரு கார் வாங்கிக் ெகாடுக்கப்பட்டது

ramaal avanukku oru car vangki-k kodukkappaddathu
(57)
ram.inst he.dat a car.nom buy.vp.san give.inf.san.pass
‘A car was bought for him by Ram.’
Further, V-V and N-V constructions can also be passivised. In such constructions,
passiviser will be suffixed to the terminal auxiliary or light verb. For instance, consider
(56) and its passive version in (57). As in causative CP constructions, passive construc-
tions can also be written as two separate words, as in example (57), or like one token -
வாங்கிக்ெகாடுக்கப்பட்டது vāṅkikkoṭukkappaṭṭatu. However, in the corpus, it was found
that the passiviser verb paṭu is always written together as one token with the light verb
as in (57) or the main verb.

3.5.5 Serial Verb Construction

Serial Verb Constructions (SVC) are common in Tamil. Unlike complex predicates, these
constructions do not satisfy the constraint of co-predication and monoclausality, following
the properties of SVCs outlined in Butt (1995), which differentiate SVC from other verbal
predicates. (58) is an example of an SVC, where both veddi ‘cut’ and vilttinaar ‘made-
fall’ share the same set of arguments and both of them are not bleached in their meaning.
Anyway, SVCs need to be investigated in much more depth.

விவசாயி மரத்ைத ெவட்டி வீழ்த்தினார்

vivasayi marattai veddi vilttinaar
(58)
farmer tree.nom cut.vpart made-fall.past.3ser
‘The farmer cut the tree down.’

29
3.5.6 Copula Construction

குமார் வக்கீல்
kumār vakkīl
(59)
Kumar.nom lawyer.nom
‘Kumar is a lawyer.’

(59) is an example of a copular construction with a nominal predicate marked with

the nominative case. The copula verb āku ‘lit. to become’ optionally occurs in nominal
predicates. āku can be used to identify the predicate when the subject and the predicate
are in the nominative case. For instance, we can conjoin āku only with vakkīl to form a
meaningful sentence in (59). The negative copula verb illai appears obligatorily to express
constitution negation as in Example (60).

குமார் வக்கீல் இல்ைல

kumār vakkīl illai
(60)
Kumar.nom lawyer.nom not
‘Kumar is not a lawyer.’

When the verb illai ‘not’ occurs as an existential negation, it is considered as the
predicate as in example (61). There are instances where the nominal predicate is dative-
case marked to express benefaction, as in example (62), and the copula is absent. In this
example āku can only be suffixed to kumārukku, meaningfully; therefore, kumārukku is
the predicate.

குமார் வீட்டில் இல்ைல

kumār vīṭṭ-il illai
(61)
Kumar.nom home.loc not
‘Kumar is not at home.’
இந்த பரிசு குமாருக்கு
inta paricu kumārukku
(62)
this gift.nom Kumar.dat
‘This gift is for Kumar’

3.6 Clitics
Clitic is a word but cannot function independently because of its dependence on an ad-
joining word. Clitics are very common in Tamil. According to the Universal Dependencies
guidelines, clitics should be separated from the main word to mark the corresponding syn-
tactic functions in the dependency structure. The only clitics found in the corpus data
are: -ā, -tāṉ, and -um, which mark an interrogative structure, the emphatic, and inclu-
siveness, respectively. For instance, examples (64)- (66) shows how interrogative marker

30
-ā is used along with the subject, the oblique noun, and the predicate to make different
questions.
ஏ ē is another clitic mentioned in the literature for marking emphasis. This will be
explored further in future versions of the communication.

அவன் ெகாழும்புக்குப் ேபானான்

avaṉ koḻumpukkup pōṉāṉ
(63)
he.nom colombo.dat.sandhi_p go.past.3sm
‘He went to Colombo.’
அவனா ெகாழும்புக்குப் ேபானான்
avaṉā koḻumpukkup pōṉāṉ
(64)
he-who colombo.dat.sandhi-p go.past.3sm
‘Is he who went to Colombo?’
அவன் ெகாழும்புக்கா ேபானான்
avaṉ koḻumpukkup pōṉāṉ
(65)
he.nom colombo.dat.-where go.past.3sm
‘Is it to Colombo he went?’
அவன் ெகாழும்புக்குப் ேபானானா
avaṉ koḻumpukkup pōṉāṉā
(66)
he.nom colombo.dat.sandhi_p go.past.3sm-did he
‘Did he go to Colombo?’

3.7 Coordination
Conjunctions in Tamil can be marked with the use of the clitic -um or the token மற்றும்
maṟṟum. While the conjoined construction using maṟṟum is straightforward, -um is am-
biguous in many cases, as it has at least eight different functions in syntax such as
in-completion, superiority, doubt, negation, completion, number, definiteness, and that
which is to come Senavaraiyar (1938).18 The conjunction works in the same way for
nouns, adjectives, adverbs and verbs. Apart from these markers, a coordinate structure
can also be marked using sociative cases on the coordinating noun phrases. When con-
joined noun phrases occur as the subject, subject-verbal predicate agreement will follow
the precedence shown below, where rationality (rat) always takes precedence.
NP1P+SG/PL+Rat + NP1P+SG/PL+Rat (+ NP3P+SG/PL+Irrat ) > VC1P+PL+Rat
NP1P+SG/PL+Rat + NP2P+SG/PL+Rat (+ NP3P+SG/PL+Irrat ) > VC1P+PL+Rat
NP2P+SG/PL+Rat + NP2P+SG/PL+Rat (+ NP3P+SG/PL+Irrat ) > VC2P+PL+Rat
NP2P+SG/PL+Rat + NP3P+SG/PL+Rat (+ NP3P+SG/PL+Irrat ) > VC2P+PL+Rat
NP3P+SG/PL+Rat + NP3P+SG/PL+Rat (+ NP3P+SG/PL+Irrat ) > VC3P+PL+Rat
NP3P+SG/PL+Irrat + NP3P+SG/PL+Irrat > VC3P+PL+Irrat
18
http://www.tamilvu.org/courses/degree/a021/a0213/html/a021331.htm

31
 Disjunction is marked with the use of the clitic ஓ ō or the token அல்லது allatu. As
in the case of coordination, this works in the same way irrespective of the disjunction
of nouns, adjectives, adverbs, and verbs. Unlike the conjunction, it is unclear how the
agreement between disjoined subjects and verbal predicates works. Since there were no
such constructions in the corpus. This will be explored more in the future version of this
communication.

3.8 Interrogatives
There are six types of interrogative structures in mentioned naṉṉūl (Thesikar, 1957).
Their choice is based on the semantic nature of the question. Syntactically, all these
questions can be constructed with the use of question particles or clitics ō and ā, or with
the use of an in-situ question pronoun, as shown in (67) and (68), respectively.

நீ ேதாைசயா ேசாறா சாப்பிடுகிறாய்

nī tōcaiyā cōṟā cāppiṭukiṟāy
(67)
you Dosa.ā Rice.ā eat.past.2s
‘Do you eat dosa or rice?’
யார் நாைளக்குக் ெகாழும்புக்குப் ேபாகிறார்
yār nāḷaikkuk koḻumpukkup pōkiṟār
(68)
who tomorrow.dat Colombo.dat go.fut.3ser
‘Who is going to Colombo tomorrow?’

3.9 Complex clauses in Tamil

Complex sentences are constructed with the use of one or more subordinate clauses. The
embedded clause or the subordinate clause is a subconstituent of the main clause. Mostly,
the subordinate clause precede the main verb of the main clause. However, since Tamil
displays free word order, a subordinate clause can also follow the main verb of the main
clause. This section briefly discusses the key types of complex sentence constructions
found in the data that are dealt with.

3.10 Non-finite clauses

Tamil has several types of non-finite clauses, including infinitives, adjectivals, adverbials,
and conditional clauses. The infinitive clause can mark the semantic interpretations of the
verb, including perception and cognitive. The infinitives are marked with -a and mostly
analysed as open clausal complement structures.

32
 நான் குமாைர ேகட்க வருகிேறன்
nāṉ kumārai kēṭk-a varukiṟēṉ
(69)
I.nom Kumar.acc ask.inf come.pres.1s
‘I come to ask Kumar.’
Adjectival clauses function like relative clauses, which will be discussed separately in
Section 3.11. Adverbial clauses are constructed using the Verb+Adverbial-marker -u that
modifies verbs, as shown in (70). These clauses are also used to construct the serial verb
construction, as shown in 3.5.5.

மாணவன் புத்தகத்ைதத் திறந்து பார்த்தான்

māṇavaṉ puttakattait tiṟantu pārttāṉ
(70)
student.nom book.acc.sandhi_t open.vpart see.past.3sm
‘The student opened the book and looked.’
Conditional clauses are formed by Verb+Conditional-marker -āl that modify verbs, as
shown in (71). This should not be confused with the -āl marker used to mark various
aspects, including the instrumental case and the agent in passive constructions.

கண்ணன் வந்தால் நான் வர மாட்ேடன்

kaṇṇaṉ vantāl nāṉ vara māṭṭ.ēṉ
(71)
Kannan.nom come.cnd I.nom come.inf will-not.1s
‘I will not come if Kannan comes.’

3.11 Relative Clauses in Tamil

Relative pronouns do not introduce relative clauses (RCs) in Tamil as in English. RCs
are formed by adding an -a morph to the verb (72), which Butt et al. (2020) refer to as
a relatives. The relative marker is null in the future participle form with -um, as shown
in Example (73).

(72) [angu nin-ṛ-a] paiyan-ai naan paar-t-en

there stand-past-rel boy-acc I.nom.1s see-past-1s
‘I saw the boy who stood there.’
(73) [angu nirk-um-∅] paiyan-ai naan paar-pp-en
there stand-fut-rel boy-acc I.nom.1s see-fut-1s
‘I will see the boy who will stand there.’

The head noun of the RC in (73) is ‘boy’, with the relative clause directly preceding
the head. One also finds RCs without a head noun in predicative contexts, as in (74).
In this case, the verb in the RC instead carries the pronominal form -atu, apart from
the relativiser -a. This -atu is form-identical with the indefinite pronoun atu, hence the
English paraphrasing with the use of ‘one’.

33
(74) [angu nin-ṛ-a-athu] en thambi
there stand-past-rel-pron.3sn my brother
‘The one who stood there is my brother.’

Example (75) involves a full head noun ‘boy’ in the accusative as the matrix object,
while (76) involves the substitution with the accusative pronoun -avan ‘he’ within the
relative clause.

(75) [angu nin-ṛ-a] paiyan-ai naan paar-t-en

there stand-past-rel boy-acc I.nom see-past-1s
‘I saw the boy who stood there.’
(76) [angu nin-ṛ-a-van-ai] naan paar-t-en
there stand-past-rel-pron.3sm-acc I.nom see-past-1s
‘I saw the one (he) stood there.’

We have done an initial study on relative clauses and reported it in Butt et al. (2020).
However, a more in-depth study on relative clauses is required, and will be covered in
future revisions.

3.12 Complimentisers in Tamil

Tamil does not have complementisers of the that-type as in English. Rather, it uses a
grammaticalised form of the verb en ‘say’, with the frozen past participle form enṛu, which
has been analysed as a type of quotative (Amritavalli, 2013; Balusu, 2020). (77) displays
the ambiguity which results between a quotative use and a complementiser function. (78)
illustrates a pure complementiser reading.

(77) ravi [naan en nanban-ai santhi-tt-en] enṛu so-nn-an

Ravi.3sm.nom [Pron.1s my friend-acc meet-past-1s] quot say-past-3sm
‘Ravi said that — “I met my friend”.’
‘Ravi said that I met my friend.’
(78) ravi [mazhai var-um enṛu] ninai-tt-aan
Ravi.3sm.nom rain come-fut.3sn comp think-past-3sm
‘Ravi thought that it will rain.’

Note that the matrix complementing verb can also take an accusative object as a
co-referent for the complementiser clause as in (79). The pure complementiser form enṛu
becomes enṛa by taking the relative marker -a. For further discussion on complementisers,
see Butt et al. (2020).

(79) [avan pizhai sey-tt-aan enṛ-a] unmaiy-ai ram nirupi-tt-aan

he mistake do-past-3sm comp-rel truth-acc Ram.nom prove-past-3sm
‘Ram proved the truth that he made mistakes.’

34
3.13 Summary
This paper outlines the key and common morphology and syntactical constructions found
in Tamil. The current constructions covered in this paper are drawn from a corpus
collected for the purpose of constructing a rule-based grammar. This corpus will be
expanded in the future, allowing for the extension of existing analyses and the inclusion
of new ones. This is an evolving document.

Acknowledgement
I extend my sincere thanks to the ZokoConnect/Herz Fellowship for their generous sup-
port during my stay at the University of Konstanz in the final quarter of 2023, which
significantly facilitated the compilation of this article. My gratitude also goes to Profes-
sor Miriam Butt and Professor Gihan Dias for their insightful guidance/input during the
paper’s preparation. Additionally, I am grateful to all my collaborators whose diverse
inputs were instrumental at various stages of my research.

References
Agesthialingom, S. (1971). A Note on Tamil Verbs. Anthropological Linguistics, 13(4):121–
125.

Alsina, A., Bresnan, J., and Sells, P. (1997). Complex predicates: Structure and theory.
In Alsina, A., Bresnan, J., and Sells, P., editors, Complex predicates, pages 1–12. CSLI
Publications, Stanford.

Amritavalli, R. (2013). Nominal and interrogative complements in Kannada. In Lasnik,

H., Park, M.-K., Miyamoto, Y., Takahashi, D., Maki, H., Ochi, M., Sugisaki, K., and
Uchiboro, A., editors, Deep insights, broad perspectives: Essays in honor of Mamoru
Saito, pages 1–21. Kaitakusha.

Annamalai, E. (2013). The Variable Relation of Verbs in Sequence in Tamil. In NINJAL

International Symposium 2013: Mysteries of Verb-Verb Complexes in Asian Languages,
pages 1–26, Japan.

Annamalai, E., Dhamotharan, A., and Ramakrishnan, A. (2014). Akarātiyiṉ putiya

patippil taṟkālat tamiḻ ilakkaṇa viḷakkam (in Tamil). In Annamalai, E., Dhamotharan,
A., and Ramakrishnan, A., editors, Cre-A: Dictionary of Contemporary Tamil, pages
xxxi–xlvii. Crea-A Publishers, India.

Arden, A. H. (1910). A Progressive Grammar of Common Tamil. The Society for Pro-
moting Christian Knowledge, Madras, India.

35
Balusu, R. (2020). The Quotative Complementizer Says “I’m too Baroque for that”. In
Formal Approaches to South Asian Languages, pages 1–12, Kansas, USA.

Boologarambai, A. (1986). A Study of Auxiliaries in the Old and the Middle Tamil. PhD
thesis, Centre of Advanced study in Linguistics, Annamalai University, India.

Britto, F. (1991). Tamil diglossia: An interpretation. Southwest Journal of Linguistics,

10(1):60–84.

Butt, M. (1994). Machine Translation and Complex Predicates. In Trost, H., editor,
Conference on Natural Language Processing (KONVENS’94), pages 62–71, University
of Vienna, Austria.

Butt, M. (1995). The Structure of Complex Predicates in Urdu. CSLI Publications,

Stanford.

Butt, M. (2010). The Light Verb Jungle: Still Hacking Away. In Amberber M, H. M.
and B, B., editors, Complex Predicates in Cross-Linguistic Perspective, pages 48–78.
Cambridge University Press, Cambridge.

Butt, M., Grimm, S., and Ahmed, T. (2006). Dative subjects. Paper presented
at NWO/DFG Workshop on Optimal Sentence Processing Available: http://ling.
uni-konstanz.de/pages/home/butt/nijmegen-hnd.pdf.Accessedon:10.08.2021.

Butt, M. and King, T. H. (2002). Urdu and the Parallel Grammar project. In Proceedings
of the 3rd Workshop on Asian Language Resources and International Standardization,
COLING, pages 39–45. Association for Computational Linguistics.

Butt, M. and Lahiri, A. (2013). Diachronic pertinacity of light verbs. Lingua, 135:7–29.

Butt, M., Rajamathangi, S., and Sarveswaran, K. (2020). Mixed Categories in Tamil via
Complex Categories. In Butt, M. and King, T. H., editors, Proceedings of the LFG09
Conference, pages 68–88, Stanford. CSLI Publications.

Caldwell, R. (1998). A Comparative Grammar of the Dravidian or South-Indian Family

of Languages. Asian Educational Services, New Delhi, India. Original work published
1875.

Chakrabarti, D., Sarma, V., and Bhattacharyya, P. (2007). Complex predicates in Indian
Language Wordnets. Lexical Resources and Evaluation Journal, 40(3-4).

Fedson, V. J. (1981). The Tamil serial or compound verb. PhD thesis, Department of
Linguistics, University of Chicago.

36
Futrell, R., Mahowald, K., and Gibson, E. (2015). Quantifying Word Order Freedom in
Dependency Corpora. In Proceedings of the Third International Conference on Depen-
dency Linguistics (Depling 2015), pages 91–100, Uppsala, Sweden. Uppsala University,
Uppsala, Sweden.

Graul, K. (1855). Outline of Tamil grammar. Leipzig: Dörfling Und Franke.

Hart, G. L. (2000). Statement on the Status of Tamil as a Classical Language. Available:

https://southasia.berkeley.edu/tamil-classesAccessedon:10.10.2018.

Irākavaiyaṅkār, M. (1958). viṉaittiripu viḷakkam (conjugation of Tamil verbs) (in Tamil).

em. Ār. Nārāyaṇaiyaṅkār, Madurai, India.

Joseph, D. (1893). A Grammar of Tamil. S.P.C.K. Press, Madras, India.

Kaplan, R. M. and Wedekind, J. (1993). Restriction and correspondence-based trans-

lation. In Proceedings of the 6th Conference of the Association for Computational
Linguistics European Chapter (EACL), pages 193–202, Utrecht University.

Krishnamurti, B. (2003). The Dravidian Languages. Cambridge University Press.

Kumar, M. A., Dhanalakshmi, V., Soman, K., and Rajendran, S. (2010). A sequence
labeling approach to morphological analyzer for Tamil language. International Journal
on Computer Science and Engineering (IJCSE), 2(6):1944–1951.

Lehmann, T. (1993). A Grammar of Modern Tamil. Pondicherry Institute of Linguistics

and Culture, India.

Lehmann, T. (1998). Old Tamil. In Stanford, B. S., editor, The Dravidian languages,
pages 75–99. Routledge London, London.

Lisker, L. (1951). Tamil verb classification. Journal of the American Oriental Society,
71(2):111–114.

Mohanan, K. P. and Mohanan, T. (1990). Dative subjects in Malayalam: Semantic

information in Syntax. In Verma, M. and Mohanan, K., editors, Experiencer Subjects
in South Asian Languages, pages 43–57. CSLI Publications, Stanford.

Mohanan, T. (1994). Argument Structure in Hindi. CSLI Publications, Stanford.

Muthuchchanmugan (2005). Ikkala Mozhiyiyal (Modern Linguistics). Mahin Press, Chen-

nai, India.

Nachinarkkiniyar (1937). Tholkappiyam - Eluththathikaram (in Tamil). Thirumahal Press,

Chunnakam, Sri Lanka.

37
Nuhman, M. (1999). aṭippaṭait tamiḻ ilakkaṇam ‘Basic Tamil Grammar’ (In Tamil).
Readers’ Association, Sri Lanka.

Pappuswamy, U. (2005). Dative Subjects in Tamil: A Computational Analysis. South

Asian Language Review, 43:XV.

Paramasivam, K. (2011). ikkālat tamiḻ ilakkaṇam ‘Contemporary Tamil Grammar’ (in

Tamil). Adaiyaalam, India.

Pope, G. U. (1979). A handbook of the Tamil language. Asian Educational Services.

Originally published: 1904.

Rajaram, S. (1986). English-Tamil Pedagogical Dictionary. Thanjavur: Tamil University.

Rajendran, S. (2004). Strategies in the formation of compound nouns in Tamil. Languages

of India, 4.

Rajendran, S. (2009). Preliminaries to the Preparation of a Spell And Grammar Checker

for Tamil. Available: https://www.academia.edu/12504639, Accessed on: 03.02.2019.

Sarveswaran, K. and Butt, M. (2019). Computational Challenges with Tamil Complex

Predicates. In Butt, M., King, T. H., and Toivonen, I., editors, Proceedings of the
LFG19 Conference, Australian National University, pages 272–292, Stanford. CSLI
Publications.

Sarveswaran, K., Dias, G., and Butt, M. (2021). ThamizhiMorph: A morphological parser
for the Tamil language. Machine Translation, 35(1):37–70.

Schiffman, H. F. (2004). The Tamil Case System. South-Indian Horizons. Felicitation

Volume for François Gros on the Occasion of his 70th Birthday, pages 293–322.

Schiffman, H. F. (2008). The Ausbau issue in the Dravidian languages: the case of
Tamil and the problem of purism. International journal of the sociology of language,
2008(191):45–63.

Senavaraiyar (1938). tolkāppiyam - eḻuttatikāram. Thirumahal Press, Chunnakam, Sri

Lanka.

Shanmugadas, A. (1982). Aspects of Tamil Language and Grammar (in Tamil). Muttamiḻ
veḷiyīṭṭuk kaḻakam, Sri Lanka.

Simons, G. F. and Fennig, C. D. (2017). Ethnologue: Languages of the world (twentieth

edition). Available: http://www.ethnologue.com Accessed on: 30.10.2018.

38
Sithiraputhiran, H. (2004). Viṉaittiripu viḷakkamum moḻiyiyal kōṭpāṭum. International
Institute of Tamil Studies.

Steever, S. B. (2005). The Tamil Auxiliary System. Routledge, New York.

Suseendirarajah, S. (1999). Studies in Sri Lankan Tamil Linguistics and Culture: Selected
Papers of Professor Suseendirarajah. Professor Swaminathan Suseendirarajah Sixty
Fifth Birthday Commemoration Volume.

Thesikar, S. N. (1957). naṉṉūl viruttiyurai (In Tamil). Vithiyanubalana Press, Chennai,

India.

39
Appendix-I: Verb forms for நட naṭa ‘walk’

நட நடந்தவைள நடப்பிக்கின்றவளால்
நடக்க நடந்தவைளக் நடப்பிக்கின்றவளுக்குக்
நடக்கக் நடந்தவைளச் நடப்பிக்கின்றவளுக்கு
நடக்கச் நடந்தவைளத் நடப்பிக்கின்றவளுக்குச்
நடக்கட்டும் நடந்தவைளப் நடப்பிக்கின்றவளுக்குத்
நடக்கத் நடந்தவள் நடப்பிக்கின்றவளுக்குப்
நடக்கப் நடந்தவனால் நடப்பிக்கின்றவளுடன்
நடக்கமுடியும் நடந்தவனுக்குக் நடப்பிக்கின்றவைள
நடக்கலாகாது நடந்தவனுக்கு நடப்பிக்கின்றவைளக்
நடக்கலாம் நடந்தவனுக்குச் நடப்பிக்கின்றவைளச்
நடக்கவில்ைல நடந்தவனுக்குத் நடப்பிக்கின்றவைளத்
நடக்கேவண்டும் நடந்தவனுக்குப் நடப்பிக்கின்றவைளப்
நடக்காத நடந்தவனுடன் நடப்பிக்கின்றவள்
நடக்காதீர்கள் நடந்தவைன நடப்பிக்கின்றவனால்
நடக்காது நடந்தவைனக் நடப்பிக்கின்றவனுக்குக்
நடக்காேத நடந்தவைனச் நடப்பிக்கின்றவனுக்கு
நடக்காமல் நடந்தவைனத் நடப்பிக்கின்றவனுக்குச்
நடக்காவிட்டால் நடந்தவைனப் நடப்பிக்கின்றவனுக்குத்
நடக்கிறது நடந்தவன் நடப்பிக்கின்றவனுக்குப்
நடக்கிறது நடந்தவார் நடப்பிக்கின்றவனுடன்
நடக்கிறார் நடந்தவார்கள் நடப்பிக்கின்றவைன
நடக்கிறவரால் நடந்தன நடப்பிக்கின்றவைனக்
நடக்கிறவருக்குக் நடந்தனள் நடப்பிக்கின்றவைனச்
நடக்கிறவருக்கு நடந்தனன் நடப்பிக்கின்றவைனத்
நடக்கிறவருக்குச் நடந்தாய் நடப்பிக்கின்றவைனப்
நடக்கிறவருக்குத் நடந்தார் நடப்பிக்கின்றவன்
நடக்கிறவருக்குப் நடந்தார்கள் நடப்பிக்கின்றவார்
நடக்கிறவருடன் நடந்தால் நடப்பிக்கின்றவார்கள்
நடக்கிறவைர நடந்தாள் நடப்பிக்கின்றன
நடக்கிறவைரக் நடந்தான் நடப்பிக்கின்றனள்
நடக்கிறவைரச் நடந்தீர் நடப்பிக்கின்றன்
நடக்கிறவைரத் நடந்தீர்கள் நடப்பிக்கின்றாய்
நடக்கிறவைரப் நடந்து நடப்பிக்கின்றார்
Continued on next page

40
நடக்கிறவர் நடந்தும் நடப்பிக்கின்றார்கள்
நடக்கிறவர்களால் நடந்ேதன் நடப்பிக்கின்றாள்
நடக்கிறவர்களுக்குக் நடந்ேதாம் நடப்பிக்கின்றான்
நடக்கிறவர்களுக்கு நடப்பது நடப்பிக்கின்றீர்
நடக்கிறவர்களுக்குச் நடப்பர் நடப்பிக்கின்றீர்கள்
நடக்கிறவர்களுக்குத் நடப்பவரால் நடப்பிக்கின்ேறன்
நடக்கிறவர்களுக்குப் நடப்பவருக்குக் நடப்பிக்கின்ேறாம்
நடக்கிறவர்களுடன் நடப்பவருக்கு நடப்பிக்கும்
நடக்கிறவர்கைள நடப்பவருக்குச் நடப்பித்த
நடக்கிறவர்கைளக் நடப்பவருக்குத் நடப்பித்தது
நடக்கிறவர்கைளச் நடப்பவருக்குப் நடப்பித்தார்
நடக்கிறவர்கைளத் நடப்பவருடன் நடப்பித்தவரால்
நடக்கிறவர்கைளப் நடப்பவைர நடப்பித்தவருக்குக்
நடக்கிறவர்கள் நடப்பவைரக் நடப்பித்தவருக்கு
நடக்கிறவளால் நடப்பவைரச் நடப்பித்தவருக்குச்
நடக்கிறவளுக்குக் நடப்பவைரத் நடப்பித்தவருக்குத்
நடக்கிறவளுக்கு நடப்பவைரப் நடப்பித்தவருக்குப்
நடக்கிறவளுக்குச் நடப்பவர் நடப்பித்தவருடன்
நடக்கிறவளுக்குத் நடப்பவர்களால் நடப்பித்தவைர
நடக்கிறவளுக்குப் நடப்பவர்களுக்குக் நடப்பித்தவைரக்
நடக்கிறவளுடன் நடப்பவர்களுக்கு நடப்பித்தவைரச்
நடக்கிறவைள நடப்பவர்களுக்குச் நடப்பித்தவைரத்
நடக்கிறவைளக் நடப்பவர்களுக்குத் நடப்பித்தவைரப்
நடக்கிறவைளச் நடப்பவர்களுக்குப் நடப்பித்தவர்
நடக்கிறவைளத் நடப்பவர்களுடன் நடப்பித்தவர்களால்
நடக்கிறவைளப் நடப்பவர்கைள நடப்பித்தவர்களுக்குக்
நடக்கிறவள் நடப்பவர்கைளக் நடப்பித்தவர்களுக்கு
நடக்கிறவனால் நடப்பவர்கைளச் நடப்பித்தவர்களுக்குச்
நடக்கிறவனுக்குக் நடப்பவர்கைளத் நடப்பித்தவர்களுக்குத்
நடக்கிறவனுக்கு நடப்பவர்கைளப் நடப்பித்தவர்களுக்குப்
நடக்கிறவனுக்குச் நடப்பவர்கள் நடப்பித்தவர்களுடன்
நடக்கிறவனுக்குத் நடப்பவளால் நடப்பித்தவர்கைள
நடக்கிறவனுக்குப் நடப்பவளுக்குக் நடப்பித்தவர்கைளக்
நடக்கிறவனுடன் நடப்பவளுக்கு நடப்பித்தவர்கைளச்
நடக்கிறவைன நடப்பவளுக்குச் நடப்பித்தவர்கைளத்
Continued on next page

41
நடக்கிறவைனக் நடப்பவளுக்குத் நடப்பித்தவர்கைளப்
நடக்கிறவைனச் நடப்பவளுக்குப் நடப்பித்தவர்கள்
நடக்கிறவைனத் நடப்பவளுடன் நடப்பித்தவளால்
நடக்கிறவைனப் நடப்பவைள நடப்பித்தவளுக்கக்கக்
நடக்கிறவன் நடப்பவைளக் நடப்பித்தவளுக்கு
நடக்கிறவார் நடப்பவைளச் நடப்பித்தவளுக்குச்
நடக்கிறவார்கள் நடப்பவைளத் நடப்பித்தவளுக்குத்
நடக்கிறன நடப்பவைளப் நடப்பித்தவளுக்குப்
நடக்கிறனள் நடப்பவள் நடப்பித்தவளுடன்
நடக்கிறனன் நடப்பவனால் நடப்பித்தவைள
நடக்கிறன் நடப்பவனுக்குக் நடப்பித்தவைளக்
நடக்கிறாய் நடப்பவனுக்கு நடப்பித்தவைளச்
நடக்கிறார் நடப்பவனுக்குச் நடப்பித்தவைளத்
நடக்கிறார்கள் நடப்பவனுக்குத் நடப்பித்தவைளப்
நடக்கிறார்கள் நடப்பவனுக்குப் நடப்பித்தவள்
நடக்கிறாள் நடப்பவனுடன் நடப்பித்தவனால்
நடக்கிறான் நடப்பவைன நடப்பித்தவனுக்குக்
நடக்கிறீர் நடப்பவைனக் நடப்பித்தவனுக்கு
நடக்கிறீர்கள் நடப்பவைனச் நடப்பித்தவனுக்குச்
நடக்கிேறன் நடப்பவைனத் நடப்பித்தவனுக்குத்
நடக்கிேறாம் நடப்பவைனப் நடப்பித்தவனுக்குப்
நடக்கின்ற நடப்பவன் நடப்பித்தவனுடன்
நடக்கின்றது நடப்பவார் நடப்பித்தவைன
நடக்கின்றது நடப்பவார்கள் நடப்பித்தவைனக்
நடக்கின்றார் நடப்பன் நடப்பித்தவைனச்
நடக்கின்றவரால் நடப்பாய் நடப்பித்தவைனத்
நடக்கின்றவருக்குக் நடப்பார் நடப்பித்தவைனப்
நடக்கின்றவருக்கு நடப்பார்கள் நடப்பித்தவன்
நடக்கின்றவருக்குச் நடப்பார்கள் நடப்பித்தவார்
நடக்கின்றவருக்குத் நடப்பாள் நடப்பித்தவார்கள்
நடக்கின்றவருக்குப் நடப்பான் நடப்பித்தன
நடக்கின்றவருடன் நடப்பிக்கிற நடப்பித்தனள்
நடக்கின்றவைர நடப்பிக்கிறது நடப்பித்தனன்
நடக்கின்றவைரக் நடப்பிக்கிறது நடப்பித்தாய்
நடக்கின்றவைரச் நடப்பிக்கிறர் நடப்பித்தார்
Continued on next page

42
நடக்கின்றவைரத் நடப்பிக்கிறவரால் நடப்பித்தார்கள்
நடக்கின்றவைரப் நடப்பிக்கிறவருக்குக் நடப்பித்தார்கள்
நடக்கின்றவர் நடப்பிக்கிறவருக்கு நடப்பித்தாள்
நடக்கின்றவர்களால் நடப்பிக்கிறவருக்குச் நடப்பித்தான்
நடக்கின்றவர்களுக்குக் நடப்பிக்கிறவருக்குத் நடப்பித்தீர்
நடக்கின்றவர்களுக்கு நடப்பிக்கிறவருக்குப் நடப்பித்தீர்கள்
நடக்கின்றவர்களுக்குச் நடப்பிக்கிறவருடன் நடப்பித்ேதன்
நடக்கின்றவர்களுக்குத் நடப்பிக்கிறவைர நடப்பித்ேதாம்
நடக்கின்றவர்களுக்குப் நடப்பிக்கிறவைரக் நடப்பிப்பது
நடக்கின்றவர்களுடன் நடப்பிக்கிறவைரச் நடப்பிப்பர்
நடக்கின்றவர்கைள நடப்பிக்கிறவைரத் நடப்பிப்பவரால்
நடக்கின்றவர்கைளக் நடப்பிக்கிறவைரப் நடப்பிப்பவருக்குக்
நடக்கின்றவர்கைளச் நடப்பிக்கிறவர் நடப்பிப்பவருக்கு
நடக்கின்றவர்கைளத் நடப்பிக்கிறவர்களால் நடப்பிப்பவருக்குச்
நடக்கின்றவர்கைளப் நடப்பிக்கிறவர்களுக்குக் நடப்பிப்பவருக்குத்
நடக்கின்றவர்கள் நடப்பிக்கிறவர்களுக்கு நடப்பிப்பவருக்குப்
நடக்கின்றவளால் நடப்பிக்கிறவர்களுக்குச் நடப்பிப்பவருடன்
நடக்கின்றவளுக்குக் நடப்பிக்கிறவர்களுக்குத் நடப்பிப்பவைர
நடக்கின்றவளுக்கு நடப்பிக்கிறவர்களுக்குப் நடப்பிப்பவைரக்
நடக்கின்றவளுக்குச் நடப்பிக்கிறவர்களுடன் நடப்பிப்பவைரச்
நடக்கின்றவளுக்குத் நடப்பிக்கிறவர்கைள நடப்பிப்பவைரத்
நடக்கின்றவளுக்குப் நடப்பிக்கிறவர்கைளக் நடப்பிப்பவைரப்
நடக்கின்றவளுடன் நடப்பிக்கிறவர்கைளச் நடப்பிப்பவர்
நடக்கின்றவைள நடப்பிக்கிறவர்கைளத் நடப்பிப்பவர்களால்
நடக்கின்றவைளக் நடப்பிக்கிறவர்கைளப் நடப்பிப்பவர்களுக்குக்
நடக்கின்றவைளச் நடப்பிக்கிறவர்கள் நடப்பிப்பவர்களுக்கு
நடக்கின்றவைளத் நடப்பிக்கிறவளால் நடப்பிப்பவர்களுக்குச்
நடக்கின்றவைளப் நடப்பிக்கிறவளுக்குக் நடப்பிப்பவர்களுக்குத்
நடக்கின்றவள் நடப்பிக்கிறவளுக்கு நடப்பிப்பவர்களுக்குப்
நடக்கின்றவனால் நடப்பிக்கிறவளுக்குச் நடப்பிப்பவர்களுடன்
நடக்கின்றவனுக்குக் நடப்பிக்கிறவளுக்குத் நடப்பிப்பவர்கைள
நடக்கின்றவனுக்கு நடப்பிக்கிறவளுக்குப் நடப்பிப்பவர்கைளக்
நடக்கின்றவனுக்குச் நடப்பிக்கிறவளுடன் நடப்பிப்பவர்கைளச்
நடக்கின்றவனுக்குத் நடப்பிக்கிறவைள நடப்பிப்பவர்கைளத்
நடக்கின்றவனுக்குப் நடப்பிக்கிறவைளக் நடப்பிப்பவர்கைளப்
Continued on next page

43
நடக்கின்றவனுடன் நடப்பிக்கிறவைளச் நடப்பிப்பவர்கள்
நடக்கின்றவைன நடப்பிக்கிறவைளத் நடப்பிப்பவளால்
நடக்கின்றவைனக் நடப்பிக்கிறவைளப் நடப்பிப்பவளுக்குக்
நடக்கின்றவைனச் நடப்பிக்கிறவள் நடப்பிப்பவளுக்கு
நடக்கின்றவைனத் நடப்பிக்கிறவனால் நடப்பிப்பவளுக்குச்
நடக்கின்றவைனப் நடப்பிக்கிறவனுக்குக் நடப்பிப்பவளுக்குத்
நடக்கின்றவன் நடப்பிக்கிறவனுக்கு நடப்பிப்பவளுக்குப்
நடக்கின்றவார் நடப்பிக்கிறவனுக்குச் நடப்பிப்பவளுடன்
நடக்கின்றவர்கள் நடப்பிக்கிறவனுக்குத் நடப்பிப்பவைள
நடக்கின்றன நடப்பிக்கிறவனுக்குப் நடப்பிப்பவைளக்
நடக்கின்றனள் நடப்பிக்கிறவனுடன் நடப்பிப்பவைளச்
நடக்கின்றனன் நடப்பிக்கிறவைன நடப்பிப்பவைளத்
நடக்கின்றன் நடப்பிக்கிறவைனக் நடப்பிப்பவைளப்
நடக்கின்றாய் நடப்பிக்கிறவைனச் நடப்பிப்பவள்
நடக்கின்றார் நடப்பிக்கிறவைனத் நடப்பிப்பவனால்
நடக்கின்றார்கள் நடப்பிக்கிறவைனப் நடப்பிப்பவனுக்குக்
நடக்கின்றார்கள் நடப்பிக்கிறவன் நடப்பிப்பவனுக்கு
நடக்கின்றாள் நடப்பிக்கிறவார் நடப்பிப்பவனுக்குச்
நடக்கின்றான் நடப்பிக்கிறவார்கள் நடப்பிப்பவனுக்குத்
நடக்கின்றீர் நடப்பிக்கிறன நடப்பிப்பவனுக்குப்
நடக்கின்றீர்கள் நடப்பிக்கிறனள் நடப்பிப்பவனுடன்
நடக்கின்ேறன் நடப்பிக்கிறனன் நடப்பிப்பவைன
நடக்கின்ேறாம் நடப்பிக்கிறன் நடப்பிப்பவைனக்
நடக்கும் நடப்பிக்கிறாய் நடப்பிப்பவைனச்
நடந்த நடப்பிக்கிறார் நடப்பிப்பவைனத்
நடந்தக்கக் நடப்பிக்கிறார்கள் நடப்பிப்பவைனப்
நடந்தது நடப்பிக்கிறார்கள் நடப்பிப்பவன்
நடந்தது நடப்பிக்கிறாள் நடப்பிப்பவார்
நடந்தார் நடப்பிக்கிறான் நடப்பிப்பவார்கள்
நடந்தவரால் நடப்பிக்கிறீர் நடப்பிப்பன்
நடந்தவருக்கக்கக் நடப்பிக்கிறீர்கள் நடப்பிப்பாய்
நடந்தவருக்கு நடப்பிக்கிேறன் நடப்பிப்பார்
நடந்தவருக்குச் நடப்பிக்கிேறாம் நடப்பிப்பார்கள்
நடந்தவருக்குத் நடப்பிக்கின்ற நடப்பிப்பாள்
நடந்தவருக்குப் நடப்பிக்கின்றது நடப்பிப்பான்
Continued on next page

44
நடந்தவருடன் நடப்பிக்கின்றர் நடப்பிப்பீர்
நடந்தவைர நடப்பிக்கின்றவரால் நடப்பிப்பீர்கள்
நடந்தவைரக் நடப்பிக்கின்றவருக்குக் நடப்பிப்பியும்
நடந்தவைரச் நடப்பிக்கின்றவருக்கு நடப்பிப்ேபன்
நடந்தவைரத் நடப்பிக்கின்றவருக்குச் நடப்பிப்ேபாம்
நடந்தவைரப் நடப்பிக்கின்றவருக்குத் நடப்பீர்
நடந்தவர் நடப்பிக்கின்றவருக்குப் நடப்பீர்கள்
நடந்தவர்களால் நடப்பிக்கின்றவருடன் நடப்ேபன்
நடந்தவர்களுக்குக் நடப்பிக்கின்றவைர நடப்ேபாம்
நடந்தவர்களுக்கு நடப்பிக்கின்றவைரக்
நடந்தவர்களுக்குச் நடப்பிக்கின்றவைரச்
நடந்தவர்களுக்குத் நடப்பிக்கின்றவைரத்
நடந்தவர்களுக்குப் நடப்பிக்கின்றவைரப்
நடந்தவர்களுடன் நடப்பிக்கின்றவர்
நடந்தவர்கைள நடப்பிக்கின்றவர்களால்
நடந்தவர்கைளக் நடப்பிக்கின்றவர்களுக்குக்
நடந்தவர்கைளச் நடப்பிக்கின்றவர்களுக்கு
நடந்தவர்கைளத் நடப்பிக்கின்றவர்களுக்குச்
நடந்தவர்கைளப் நடப்பிக்கின்றவர்களுக்குத்
நடந்தவர்கள் நடப்பிக்கின்றவர்களுக்குப்
நடந்தவளால் நடப்பிக்கின்றவர்களுடன்
நடந்தவளுக்குக் நடப்பிக்கின்றவர்கைள
நடந்தவளுக்கு நடப்பிக்கின்றவர்கைளக்
நடந்தவளுக்குச் நடப்பிக்கின்றவர்கைளச்
நடந்தவளுக்குத் நடப்பிக்கின்றவர்கைளத்
நடந்தவளுக்குப் நடப்பிக்கின்றவர்கைளப்
நடந்தவளுடன் நடப்பிக்கின்றவர்கள்

45