Lomashishiksha (Lomaśīśikṣā)

The Lomaśī Śikṣā of the Sāmaveda.
I shall explain the Lomaśī [Śikṣā] conceived by the teacher Garga,
complete with names, just as it has been taught and according to the words of one's teacher.

The short, the long, and likewise the protracted, sung out by the Sāmavedic chanters,
they should execute repeatedly like a mṛdaṅga drum, like the signs of falling down and rising up.

By what is the shake surpassed by a shake? Through what does a conjunct shake?
Or what is called a shake, by which it is designated as a shake?

Struck by the preceding member beforehand, but held by the succeeding member,
split into two by a consonant, the frightened vowel shakes.

In a long shake, one and a half morae are prescribed on the third [tone];
but on the second, half a mora: the practice is like the striking of a bell.

Nasalisation (raṅga) is to be known as two-fold: having a following vowel or having a following consonant.
Having the appearance of the same colour as a pigeon, it is prescribed by scholars of syllables.

Its mora is in the heart, but an atomic measure is in the head;
at the tip of the nose, indeed, a measure of atoms of nasalisation is proclaimed.

And when nasalisation has arisen, one should not swallow the preceding syllable;
on the vowel, one should employ a long [measure] and pronounce its nasal quality.

Having two morae or having one mora, relying on the base of the nose,
nasalisation is applied at the end with the fifth [nasal consonants], entirely nasal.

Thus ends the first section.

Half of the sound k and half of the sound ṭ remain in the right jaw;
one should pronounce them with the tip of the tongue in the form of 'kiṭ-kīḍā'.

Kareṇu, Kurviṇī, Hariṇī, Hāritā,
and also Haṃsapadā by name: these are the five svarabhaktis (anaptyctic vowels).

One should know Kareṇu in [the combination of] t and s, Kurviṇī in ṛ and dh or k;
one should know Hariṇī in d and ṛ or ś, and Hāritā in ṛ and ś;

and one should pronounce Haṃsapadā when r is followed by ṣ;
that which occurs between r and ṣ, one should designate as Kākinī.

The learned always desire [gemination] after r and h;
but where h and r are seen [together], no doubling occurs for the two.

If h is seen intermediate between two r-sounds anywhere,
one should recognise gemination there; 'preṅkhāṃstarhye' is the example.

Thus ends the second section.

It does not sound on the preceding member, nor ever on the succeeding member,
neither on the consonant nor on the mora; how then is the junction prescribed?

On the prior half of the vowel itself, and on the latter quarter of the consonant;
upon the junction of their halves, the wise person should produce the tone.

When a word is in doubt whether it is accented on the end, accented in the middle, or paused,
one should recognise it as accented on the end if the suffix is acute (udātta).

When a word is in doubt whether it is accented on the end, low-circumflexed, or paused,
one should recognise it as accented on the end if the suffix is acute (udātta).

Throwing down (avakṣepa), measure, noise, and also distress,
doubling (yamaka), and circumflex: these five are remembered as faults in the Ṛgvedic collection.

The speech-sounds whose gemination is desired are twenty-one:
namely, the fifth [nasals], the third [voiced stops], y, l, v, together with ś, ṣ, and s.

In an anusvāra, in hiatus (vivṛti), in pause, and in a pair of identical vowels,
one should bring out both lips; likewise in the case of o and y.

Thus ends the third section.

Even of many u-sounds, a single articulatory effort is prescribed;
in the short, the long, and the protracted, the primary articulatory effort should not be diverse.

Even among many u-sounds, if the middle one is long,
the articulatory effort is at the beginning and the end, but the middle one is separate.

Even among many u-sounds, if the final one is long,
the articulatory effort is at the beginning and the end, while the middle one relies on a separate [effort].

Where there is a double articulatory effort at the beginning and afterwards the sound u,
that is to be known as beginning, middle, and end; 'vāyo' [and] 'uktaḥ' is the example.

Where there is a single articulatory effort at the beginning and afterwards the sound u,
that is to be known as beginning and end; 'vāyo śukraḥ' is the example.

If the preceding sound is non-conjunct and the following sound is subsequent,
and it is conjoined with an u-sound, there should be a repetition of articulatory effort.

The consonants following a vowel, and those suitable at the end of an u-sound—
one should know the articulatory effort here of these, and then of o and ḍ.

Wherever two contiguous labial vowels are joined in conjunction,
there should be a single articulatory effort; otherwise the articulatory effort is doubled.

If two sounds are labials, even when followed by a non-conjunct,
one should produce double labiality with them; 'ojiṣṭhaṃ pupu-' is the example.

Thus ends the fourth section.

What follows from a preceding word ending in a vowel changes into ṣ or ṇ;
when your word ceases in the verses of the Sāmaveda, there is recommencement like a junction (sandhi).

n and m are heavy, termed resonance (nāda); light, they are nasal;
when conjoined and when accompanied by visarga, they are indeed proclaimed as resonance.

In hiatus (vivṛtti), in pause, and when a homogeneous suffix follows,
they are to be known as nasal; the rest are proclaimed as resonance.

The nasal is known to have one mora, those called resonance have two morae;
those with visarga are resonance, and those with anusvāra are as before.

Those sounds which are nasal, if voiced sounds follow as suffixes,
one should recognise resonance there; 'anudha' is the example.

Even among nasal sounds, whatever tone is assigned where,
they conform to precisely that; and one should avoid faults of the nose.

One atom on the thumb, and likewise on the forefinger;
and on the third to be made, the inclined [tone] is thus applied.

Two atoms within the thumb, six are remembered on the forefinger;
half a mora on the third, the inclined [tone] is thus applied.

The sound h conjoined with the fifth [nasal consonants] or conjoined with the semivowels,
that one should recognise as pectoral (produced in the chest); unconjoined, they call it guttural (produced in the throat).

The sound h conjoined with r, and sometimes joined with v, l, or s,
is to be pronounced in the throat; thus the learned desire it.

The sound h which stands prior, and following which is dy—
if separated untimely in continuous recitation (saṃhitā), it is not pectoral.

Thus ends the fifth section.

The consonant r placed atop another consonant enters into the nature of the vowel ṛ;
but that is not to be done everywhere: it is to be done before ś, ṣ, s, and h.

Where a syllable and ś, ṣ, s stand conjoined,
one should know the vowel ṛ there, but designate it as a consonant.

Where a syllable is seen as a consonant at the pause of a word,
one should recognise that as the preceding member; that which stands [after] is like the succeeding member.

Of the r-sound and of the vowel ṛ, and likewise of the svarabhakti,
the application is like the sound a, and like a double consonant.

Reaching the intermediate pitch (śruti), one should show the two-fold tone:
one should do it like 'syeniṣi' for the downward [tone], and 'brahma jajñānam' when going upward.

From having six limbs it is inclined (praṇata); it is repeatedly called low (nīca);
'āyāhi' is inclined and low in 'pariprāṇa' and on the fourth.

When that which ends in the sound ā [and] ends in a sibilant is followed by a voiced sound, / there is elision before the voiced sound; such is seen in vidāgādha.

Thus the sixth section.

In astha, ustha, and in nasalised resonance (raṅga), and likewise upon the elision of a gati, / and upon the elision of e-kāra and o-kāra, there is hiatus (vivṛtti) followed by a consonant.

Non-suffixal u-vowels followed by dha, ḍha, or ta, except in pakvi, / are understood as closed (saṃvṛtta) among u-vowels, except in the case of harīta.

That [pause] which ceases in the second [svara] does not proceed in the fourth; / one should know the bright [intonation] upon the arrival of agho, but above it is lengthened (āyatī).

Ūtā, prahū, masā, and also tsivā, vārā, likewise mahī— / the wise direct that these are always to be pronounced lengthened.

In prāṇāpāna, nigāra, saṃsvāda, and also in ṛṣabha, / they are to be performed on the first tone, and also kīṭi and kaḍākāra.

One should release the breath continually; the breath should first be through the right [nostril]; / but inhalation of breath should be by the left, or with high pitch.

The breaths are released through the right [nostril]; through lack of measure it would be otherwise; / having drawn in by the left, such is considered to be the tone of one who is immeasurable.

Those u-vowels preceded by a pāda, and likewise those preceded by a pada, / should be understood as ḍa and i; this rule applies also in the case of ḍa and ḍha.

Just as fresh butter is produced when curds are churned, / in the same way, the desired [sound] below those two should be realised when joined separately.

Where śa, ṣa, sa, and ha appear, following upon them, / one should recognise a long anusvāra, or a short one in a conjunct.

After a non-conjunct, the anusvāra attains the state of light [syllables]; / but after a conjunct it is always [heavy], and doubt ceases.

When a conjunct follows and the anusvāra is preceding, / one should know that [preceding vowel/mātrā] to be short; 'saṃsthe' is the illustration.

If a closed [sound] is seen following the anusvāra, / that is understood to be long; hear that twofold [rule] of ours.

The anusvāra has two morae beginning from one mora, and from two morae it has one mora; / even from a single mora in a conjunct, it has a single mora in a twofold manner.

Thus the seventh section.

When ya and va are joined with r, if the r is first, / one should understand it as belonging to the preceding limb; 'sūrya' is an example of the preceding.

When ya and va are joined with r, and if it occurs at a pause, / one should understand it as belonging to the following limb; 'aro harya' is the defining instance.

From āsannāt, mahānāt, and yātudhānāt, as they occur, / the wise teach that these do not take a preceding n-sound.

Wherever the letter h is seen joined with the class-nasals, / they are understood as beginning with ut, and one should realise them joined separately.

A letter h preceded by r, or even following after r, / [or] where h follows an anusvāra—it duplicates (kramati) in these three [positions].

The sounds śa, ṣa, and sa following an anusvāra never duplicate; / śa, ṣa, and sa duplicate there if followed by semivowels.

Where a speck resembling a ray of the sun is seen, / that is the measure of an atom (aṇu); but a mora (mātrā) consists of four atoms.

One should know [the sound] residing in the heart as an atom; in the throat, one should know it as two atoms; / as three atoms at the tip of the tongue; released [forth], it becomes one mora.

The blue jay utters one mora; the crow has declared two morae; / the peacock is to be known as having three morae: this is the comprehension of morae.

A short [vowel] has one mora; a long [vowel] is said to have two morae; / a prolated [vowel] is understood as having three morae; and a consonant has half a mora.

Everything here is a fourth; they do not increase; these increase: / as in daścarati, rāndrasvindrasacchū, atapasātvāsūgolārātpatantī, and in the secret text / of varhirṣya etc., there is non-circumflexion, elision of the consonant, elision of the consonant.

Thus the Lomaśī Śikṣā is completed.
In the year joined with planets (9), Vedas (4), and limbs (6) [1649 VS], in the month of Iṣa (Āśvina), in the bright fortnight, / the Gautamī and Lomaśī Śikṣās were prepared for the delight of the Sāmavedic chant-singers,

transcribed at the Veda Vidyālaya for the welfare of the Sāmaveda chanters / by the compassionate reciter Yugalaprākkiśora.