< All linguistic notions

Articulatory Phonetics in Ancient India

Phonetics is the branch of linguistics that studies the physical properties of speech sounds. Within phonetics there are three sub-branches: articulatory, which focuses on the anatomy and physics of sound generation; acoustic, which examines the acoustic properties of the sounds of the world's languages; and auditory, which studies the features of the signal that listeners use to perceive the message (Whalen 2019). Among these three branches, modern articulatory phonetics was greatly influenced by the phonetic categories and terminology from ancient Indian linguistics.

Articulatory phonetics is concerned with describing the speech sounds of the world's languages in terms of their articulations, that is, the movements and/or positions of the vocal organs (articulators). The most influential system of articulatory description and transcription of speech sounds is the International Phonetic Alphabet (IPA), which aims to provide a phonetic symbol for every distinct sound for every language (Whalen 2019). These sounds are defined as a combination of articulatory properties and grouped as different segment types (i.e. consonants and vowels).

In ancient India, a relatively sophisticated analysis of articulatory phonetics developed at an early stage as part of attempts to ensure the accurate oral transmission and memorization of early Vedic texts. Early references to an inventory of sounds (akṣarasamāmnāya) can be found in the Aitareya Āraṇyaka (3.2.1) and the Chāndogya Upanisạd (2.22.3-5) (Lowe 2020). The earliest texts dedicated to phonetic descriptions were called śikṣās 'instruction'. Initially, the term śikṣā seems to have been restricted to rudimentary instruction in pronunciation, which included instruction in individual sounds, accent, quantity, and chanting of Vedic verses (e.g. in the Taittirīya Upaniṣad, Varma 1961: 4). Later on, the term was used to refer to “general phonetics”. Most of the surviving śikṣā texts are of a relatively late period; the best known is the Pāṇinı̄yaśikṣā, attributed by the tradition to the famous Sanskrit grammarian Pāṇini (who lived ca. 400 B.C.). Other important śikṣās include the Vyāsaśikṣā, the Āpiśaliśikṣā, the Yājñavalkyaśikṣā, and the Nāradaśikṣā (Deshpande 2023).

We also find the prātiśākhyas (prati 'each' + śākhā 'branch'), or pārṣadas, which present phonological descriptions from the four Vedas. They describe the correct pronunciation based on two modes of recitation: saṃhitāpāṭha (i.e. running text) and padapāṭha (i.e. words separated and pronounced in pausa). The main concern of the prātiśākhyas was accounting for sandhi (lit. 'joining together') phenomena between words, but they also address pronunciation and other matters.

Ancient Indian phoneticians carefully described the sounds (varṇa) of Sanskrit and grouped them into different categories, distinguishing mainly between consonants (vyañjana), vowels (svara), semivowels (antasthā), and fricatives (ūṣman). The following sections present the relevant terminology used to describe these categories in the ancient Indian tradition.

1. Karaṇa 'articulator'

The articulators (karaṇa) are mainly the different parts of the tongue involved in the articulation of a sound. According to the Āpiśaliśikṣā, for dental, retroflex, palatal, and velar sounds, the articulator is the tongue, particularly: the tip of the tongue (jihvāgra), almost the tip of the tongue (jihvopāgra), the underside of the tip of the tongue (jihvāgrādha), middle of the tongue (jihvāmadhya) and the root of the tongue (jihvāmūla). We also find the mention of jihvāntābhyām 'with both sides of the tongue'. In other instances, the articulator is the same as the point of articulation (e.g. throat, nasal, the lower lip in a labial sound).

2. Sthāna 'point of articulation'

The points of articulation (sthāna) are: kaṇṭha 'throat', tālu 'hard-palate', mūrdhan 'top of palate', danta 'teeth', dantamūla 'alveolar ridge', oṣṭha 'lips', nāsikā 'nose', uras 'chest'.

Velar consonants (kaṇṭhya) include k, kh, g, gh, ṅ. This series is called ka-varga and it is generally described as being produced with the velar (throat) point of articulation and the articulator jihvāmūla 'root of the tongue'. The nasal final sound () additionally involves the nāsikā 'nose'. The sounds h (hakāra) and (visarga) are also grouped in this series, but some consider as produced in the chest (uras), according to the Āpiśaliśikṣā.

The term kaṇṭhya is also used to describe the simple (samāna) vowels a, ā. The compound vowels (sandhyakṣaras), e/ai and o/au, are described with a combination of points, as velar-palatal and velar-labial, respectively.

Palatal consonants (tālavya) include the sounds c, ch, j, jh, ñ. These sounds are described as being made with the middle of the tongue (jihvāmadhya) against the palate. The sibilant sound ś is also described as palatal. The simple vowels i, ī, as well as the semivowel y, are also classified as palatal.

Retroflex consonants (mūrdhanya) include the sounds ṭ, ṭh, ḍ, ḍh, ṇ. The term prativeṣṭita 'rolled back' is used in the texts to refer to these sounds, along with the term mūrdhan meaning 'top of the palate'. The articulator is the part next to the tip of the tongue (jihvopāgra) which is retroflexed. The sibilant sound is also described as retroflex.

The semivowel r, as well as vocalic , , are also considered retroflex, but according to the Āpiśaliśikṣā, some consider r as dental and some as alveolar (dantamūlīya 'root of the teeth'). Another feature of the r is repha 'snarl'.

Dental consonants (dantya) include t, th, d, dh, n. The articulator is described as the tip of the tongue (jihvāgra), which is further described as prastīrṇa 'spread, flat'. The sound l is generally treated as a semivowel associated with the dental class, and it is produced using the side (anta) of the tongue. In this series, we also find the sibilant sound s.

The vocalic (and for some traditions) is also classified as dental, although there are also references to alveolar articulation. This is also described as involving both sides of the tongue (jihvāntābhyām).

Labial consonants (oṣṭhya) include the sounds p, ph, b, bh, m. Here the upper lip is often referred to as the place of articulation and the lower lip as the articulator. This place of articulation also includes the semivowel v, which is described as dental-labial. The simple vowel u/ū is classified also by the labial feature, rather than the tongue position, the shape of the lips is referred to as oṣṭhopasaṃhara 'contracting of the lips' or as oṣṭhau dīrghau 'lengthening of the lips'.

3. Prayatna 'effort, manner of articulation'

How the physical organ or karaṇa 'articulator' relates to the point of articulation is called prayatna 'effort or manner of articulation'. These processes are divided into two main types: ābhyantara 'internal' and bāhya 'external'. The first type comprises processes occurring within the buccal cavity (i.e. 'intra-buccal') and the second those occurring elsewhere (i.e. 'extra-buccal'). Intra-buccal processes (ābhyantara-prayatna) include: closure, opening, and constriction. Extra-buccal processes (bāhya-prayatna) include: glottal, pulmonic, and nasal. These extra-buccal processes are discussed in section 4 and 5.

Sounds are divided according to the degree of closure between sthāna and karaṇa in four types:

  1. sparśa 'contact sounds' (stops and nasals),
  2. antasthā 'in-between sounds' (semi-vowels),
  3. ūṣman 'aspiration sounds' (spirants),
  4. and svara 'vowels'.

There is the notion of minimal closure or 'non-contact' (aspṛṣṭa) that distinguishes vowels from consonants. Among vowels, the aperture is the widest with ā, least with i, u, ṛ, ḷ, according to the Āpiśaliśikṣā. In some texts, there is a mention of the 'contracting of the lips' for the articulation of u/ū.

On the other hand, maximal closure describes the stops (sparśa). The Āpiśaliśikṣā further refers to four intermediate degrees:

  1. spṛṣṭa 'contact'
  2. īṣat-spṛṣṭa 'slight contact'
  3. īṣad-vivṛta 'slight openness'
  4. vivṛta 'openness'

The term īṣad-vivṛta 'slight openness' is used to describe fricatives (Allen 1953: 25). Note that fricatives are articulated in the same places as the corresponding stops, differing only in closure. The term used for fricatives is ūṣman 'steam, vapour'. This term is applied also to h, and to the breathy release of the aspirated stops (soṣman). According to Allen (1953), there is no special term corresponding to 'sibilant', though excessive sibilation is referred to as kṣveḍanam 'whistling'.

The term īṣat-spṛṣṭa 'slight-contact' is used to describe semivowels (antasthā). It is interesting to note that there is a greater degree of contact of this class than in fricatives. In the śikṣās it was stated that y and v in initial position involve greater contact than the vowels i/ī and u/ū.

4. Anupradāna 'phonation'

As part of the extra-buccal processes (bāhya-prayatna) there is the distinction based on whether the glottis is open or closed. This is called anupradāna 'phonation'. The tradition recognizes two types: śvāsa 'breath' and nāda 'resonance', resulting from two different positions of the glottal chords: vivṛta 'open' and saṃvṛta 'closed'. The first type (śvāsa) further results in voiceless sounds, whereas the second type (nāda) in vowels and voiced consonants.

Interestingly, texts like the Taittirı̄ya-Prātiśākhya also include an intermediate condition between closed and open, where both breath and voice are produced (Deshpande 2023: 85). This is the case of the voiced fricative h and the voiced aspirates. Later grammatical and phonetic terminology made the distinction between aspirate and non-aspirate, namely mahāprāṇa 'with greater prāṇa' and alpaprāṇa 'with less prāṇa', respectively.

So, we have voiced aspirated/non-aspirated and voiceless aspirated/non-aspirated. According to the Pāṇinīyaśikṣā, h and the voiced aspirates are voiced, the semivowels and voiced stops are partly voiced; the voiceless aspirates are breathed, the voiceless stops are partly breathed (Allen 1953: 37-38). Furthermore, the difference between h and is that the first one has the features breath + voice, whereas the second one is 'pure breath'. Furthermore, the sounds (jihvāmūlīya) and (upadhmānīya) are considered variants for (Deshpande 2023: 84).

The Western tradition did not understand the articulatory processes involved in voicing until it came into contact with the Indian tradition. The Indian tradition uses two terms to refer to voicing: nāda and ghoṣa. There is a general understanding that nāda is shared by vowels and ghoṣa by consonants (Deshpande 2023, 1976). Deshpande (1976) has suggested that there is an articulatory difference between the two, whereas Cardona (1986) has suggested that the difference is only phonologically motivated. The voiced phonemes include semivowels, h and anusvāra ().

Chatterji (1948: 239) argues that the sense of ghoṣa was shifted from 'vowels' to 'voiced consonants' from the early prātīśākhyas. In the prātīśākhyas and śikṣās, the term ghoṣavat is consistently applied to voiced consonants alone, and not to vowels. Another possible analysis this author presents is that the terms śvāsa and nāda refer to the speaker and ghoṣa and aghoṣa refer to the hearer; in other words, ghoṣa and aghoṣa are acoustic effects of śvāsa and nāda, respectively.

5. Nāsikya 'nasal', anunāsika 'having a nasal component', and anusvāra ()

We find nasal consonants in all the series described in section 1, these are called nāsikya 'nasal'. But we also find nasalized forms of three of the semivowels (ỹ, l̃, ṽ).

Regarding vowels, we find the term anunāsika which is used in opposition to the term śuddha 'pure non-nasalized vowels'. The anunāsika is a nasal sound lacking the closure (vivṛta) which is required to produce a nasal stop. It is described as being produced by the nose and the mouth. Elsewhere, the term can be used for nasal consonants.

According to Allen (1953: 40), some texts also use the term rakta 'coloured', where nasalization is being referred to as a rāga or raṅga, i.e. 'nasal colour'. The nasalized vowels are not of frequent occurrence, since they appear in certain types of junction. Apart from such cases there is a tendency for vowels to take some degree of nasal 'colour' in contact with nasal consonants.

There is also a third type called anusvāra () 'after-sound or subordinate sound'. This is restricted to post-vocalic position, and its primary context is before the fricatives (ś, ṣ, s and h), in cases where historical and phonological evidence point to an alternation with m or n, and at an early date also before r. The anusvāra is also a nasal sound lacking the closure (vivṛta) which is required to produce a nasal stop, and it is described as being uttered only through the nose.

6. Svara 'tone'

The Indian tradition recognizes the existence of three tone classes in Vedic: udātta 'high tone', anudātta 'low tone', and svarita 'falling'. Note that the first half of svarita is udātta, which falls in the second half (i.e. ākṣipta 'cast down' or pravaṇa 'downhill slope'). In the Pāriśikṣā, udātta is described as produced by tension, hardness and constriction of the glottis; anudātta is produced by laxness, softness and widening of the glottis (Allen 1953: 90).

7. Yamas 'twin, transitional sounds'

We find in the Indian tradition a detailed description regarding the transition between the stop and nasal. There are also the so-called transitional sounds or yamas which are nasal (nāsikāmūlīya 'produced at the root of the nose') and produced between a stop and a following nasal. There is a slight nasalization of the oral stop, at least in its latter portion, and so a nasal off-glide to the following nasal.

8. Summary

In the Indian tradition, as in the IPA, consonants and vowels are described as a combination of articulatory features that involve the point of articulation, the articulator, and the manner of articulation. Consonants are grouped in vargas according to their point of articulation, and they are described in terms of voicing, aspiration, and the degree of closure or opening between the articulator and the point of articulation. Semivowels are considered to present more closure than vowels, and even more than fricatives. Vowels, on the other hand, are described primarily based on their point of articulation, with reference also to degree of closure (corresponding to vowel height); in contrast with modern phonetics there is no discussion of frontness or backness; roundedness of the lips is mentioned only in reference to u/ū. Also note that the ordering of the points of articulation differs from the Western system, starting with the velar sounds and moving forwards.



Sanskrit Words

Text Passages