V2a

Antiquity of the Śivasūtras: Phylogenetic Constraints on the Vowel Inventory

Abstract

The phonological distinctions in the vowel section of the Śivasūtras presuppose the capacity for quantal vowel production, while indicating articulatory capabilities distinct from those of anatomically modern humans. This constrains the composition of the Śivasūtras to a period following the evolution of biological capacities for quantal vowels, yet preceding the extinction of non-sapiens Homo species (~40 kya). A more precise lower bound emerges from the absence of certain abilities in Homo sapiens, placing the vowel inventory’s origin prior to the emergence of modern human anatomy.

Introduction

In Sanskrit grammar, all words are derived forms. Pāṇini’s Aṣṭādhyāyī comprises approximately 4,000 sūtras describing word derivation, reliant on a foundational list of phonemes (varṇas) known as the Māheśvara Sūtras or Śivasūtras. This phonemic inventory underpins Pāṇinian grammar (Aṣṭādhyāyī and Pāṇinīya Śikṣā) and represents one of the earliest preserved linguistic artifacts. The present analysis examines the antiquity of the Śivasūtras’ vowel (svara) section through phylogenetic dating, integrating articulatory phonetics, comparative primatology, and paleoanthropology.

The Śivasūtras divide into vowels (svaras) and consonants (vyañjanas). The consonant section postdates the Hominini divergence (~3.1–3.3 mya [Ma]), as extant non-human Hominini (Pan troglodytes, Pan paniscus) lack full consonant production capabilities.

Chronological bounds for the Śivasūtras are established as follows:

1. Upper bound (terminus post quem): Determined by the sequential emergence of individual svaras in the hominin lineage, with the composite inventory feasible only after all svaras were producible.

2. Lower bound (terminus ante quem): Inferred from (a) svaras present in the Śivasūtras but extinct in Homo sapiens per Pāṇinīya Śikṣā descriptions, and (b) svaras producible by Homo sapiens but absent in the Śivasūtras. This yields a pre-sapiens origin, pending vocal tract reconstructions of Homo erectus and earlier taxa.

Key phonological features include:

• A 2 + 5 + 2 svara system
• Presence of अ, इ, उ; absence of आ
• ऋ and ऌ as svaras and a-spṛśya (non-contact)
• Dual forms of ह
• Absence of a 3 × 3 × 2 tonal/length/nasal grid
• Absence of ळ

The 2 + 5 + 2 Svara System and Its Antiquity (Upper Bound)

The vowel inventory in the Śivasūtras comprises three categories—growl vowels (ऋ, ऌ), quantal vowels (अ, इ, उ, ए, ओ), and derived vowels (ऐ, औ)—whose production requires specific articulatory and acoustic mechanisms traceable through hominin and primate evolution. Phylogenetic mapping of these “characters” (presence/absence of vowel types and combinatory abilities) constrains the upper bound for the vowel system’s compilation to the emergence of these capacities in the last common ancestor with relevant taxa.

• Quantal vowels (अ, इ, उ, ए, ओ): These five vowels exhibit stable formant patterns (F1/F2 dispersion-focalization) that minimize acoustic-perceptual confusion, as modeled by Stevens (1989) and quantified in human supralaryngeal vocal tract (SVT) simulations (Lieberman et al., 1969; Lieberman, 2012). Acoustic analyses of Guinea baboon (Papio papio) vocalizations reveal homologous vowel-like segments (VLSs) matching human [ɨ æ ɑ ɔ u], organized along two axes: horizontal (tongue advancement: [æ] ⇔ [u ɔ]) and vertical (tongue height: [ɑ] ⇔ [ɨ]) (Boë et al., 2017). Baboons achieve this despite a high larynx, using tongue musculature (genioglossus, hyoglossus, styloglossus) comparable to early hominins. This proto-system dates to the last common ancestor with Cercopithecoidea (~25 Ma), as baboon VLSs falsify the high-larynx barrier hypothesis and indicate pre-hominin articulatory precursors (Boë et al., 2017). Thus, the capacity for these five vowels imposes a terminus post quem of ~25 Ma for the Śivasūtras’ quantal set.

• Derived vowels (ऐ, औ): These arise from diphthongization (अ+इ → ऐ; आ+ओ → औ), requiring combinatorial sequencing. Comparative data show this ability predates quantal vowels: chimpanzees concatenate calls for context-specific meaning (Girard-Buttoz et al., 2025), Campbell’s monkeys form sequences (Ouattara et al., 2009), suricates vary alarm calls acoustically (Manser, 2001), and cetaceans/birds combine units (Payne & McVay, 1971). Since baboon “wahoo” sequences ([æ]-[ɔ]) demonstrate succession ~25 Ma (Boë et al., 2017), ऐ/औ compatibility aligns with the same bound.

• Growl vowels (ऋ, ऌ): ऋ and ऌ are described in the Pāṇinīya Śikṣā as non-contact (a-spṛśya) sounds, produced without any contact (sparśa) of the tongue, teeth, or lips with other articulators (Pāṇinīya Śikṣā 4). Producing ऋ and ऌ in this Śikṣā-defined manner—as stable, contrastive, and combinable svar—is biomechanically incompatible with the vocal tract control of anatomically modern humans. As an illustration of the intended vocal gesture, partial acoustic approximations can be produced by externally restricting lingual mobility (e.g., gently holding the tongue) while attempting to articulate ऋ and ऌ, yielding sounds resembling a roar (ऋ) and a yelp or wail (ऌ), respectively.

The non-contact (a-spṛśya) production of ऋ and ऌ, described in Pāṇinīya Śikṣā as sounds generated without any tongue, teeth, or lip touch/contact (sparśa) with other articulators, corresponds to motivation-structural rules (MSR): ऋ approximates a low-frequency, harsh roar associated with dominance, superiority, or threat displays, while ऌ approximates a high-frequency yelp or wail signaling submission, appeasement, or fear (Morton, 1977). These MSR patterns—where acoustic structure reflects motivational state—are phylogenetically ancient. It must have been established in auditory/vocal form prior to the synapsid-sauropsid divergence (~310–330 Ma); the ubiquitous presence of these specific frequency-to-motivation correlations across both extant mammals and birds confirms that the last common ancestor of all amniotes utilized this acoustic signaling system for social communication (Morton, 1977; Benton & Donoghue, 2007). Gestural analogues of such aggressive-submissive signaling appear in the elaborate waving displays of fiddler crabs (genus Uca), where lateral claw-waving serves threat (aggressive/superiority) or courtship/submissive functions, representing one of the earliest documented forms of visual signaling for dominance and submission in bilaterians, phylogenetically traceable to the common ancestor of arthropods and deuterostomes (>540 Ma; Crane, 1975). Neural circuitry for interpreting superiority-submission signals predates audition/phonation.

The discussion of Motivation–Structural Rules is not intended to derive the phonological inventory itself, which lies beyond the scope of the present work, but rather to situate the emergence of meaningful vocal distinctions within a deeper pre-linguistic communicative continuum.

The combined vowel system thus requires anatomical/neural infrastructure present by ~25 Ma (upper bound from cercopithecoid divergence), as evidenced by baboon acoustics and primate combinatorics. This establishes the biological substrate enabling the Śivasūtras’ inventory, without implying compilation at that epoch.

Absence of आ (Lower Bound)

The presence of अ, इ, उ, ए, ओ establishes an upper bound of ~25 Ma, as detailed above. The absence of ā provides a lower bound. Traditional Sanskrit grammar treats dīrgha a as ā, yet Aṣṭādhyāyī 8.4.68 distinguishes them (Pāṇini, n.d.). ā appears in the Aṣṭādhyāyī but not the Śivasūtras, constraining the latter to predate distinct ā production, and the former to postdate it. The Aṣṭādhyāyī compiles inherited knowledge; thus, it either originated post-ā or was recompiled thereafter. In either case, the Śivasūtras predate ā capability, hence predate Homo sapiens.

ऋ and ऌ as Svaras and A-spṛśya

Pāṇinīya Śikṣā classifies ऋ and ऌ as svaras distinct from spṛśya (contact) sounds (Pāṇinīya Śikṣā 4). Produced without intraoral contact, they yield roar (ऋ) and yelp/wail (ऌ) forms. Gestural equivalents trace to ~500–550 Ma (Crane, 1975); acoustic forms to ~310–330 Ma (Morton, 1977). Homo sapiens cannot produce these per Śikṣā prescriptions. Given ā absence and ऋ/ऌ presence, the Śivasūtras must predate Homo sapiens.

The growl vowels ऋ and ऌ may be associated with laryngeal air sacs, which were present in early Homo until approximately 600 ka (de Boer, 2012). Their extinction provides a potential terminus ante quem of ~600 ka for these sounds. This is consistent with evidence of seafaring capability in Homo erectus at ~1.0–1.5 Ma (Hakin et al., 2025), which would imply the presence of structured language.

The association between motivation-structural rules (MSR) in vocal communication and laryngeal air sacs in the Homo lineage remains unstudied, as extant Homo sapiens lacks laryngeal air sacs (de Boer, 2012). Consequently, the proposed terminus ante quem of ~600 ka for the retention of MSR-linked growl vowels (ऋ, ऌ) cannot be established with certainty and requires further paleoanthropological and comparative evidence.

Two Types of ह

The Śivasūtras distinguish two ह forms, produced differently yet auditorily indistinguishable in Homo sapiens. This retention of perceptual distinction indicates pre-sapiens origin.

Absence of 3 × 3 × 2 Grid

Sanskrit svaras exhibit 18 varieties: three tones (udātta, anudātta, svarita), three durations (hrasva, dīrgha, pluta), and nasal/non-nasal (3 × 3 × 2). Absent in the Śivasūtras, this may reflect brevity or absent fine motor control. Present in Homo sapiens, absence implies pre-sapiens composition (~40–315 ka).

Absence of ळ

ळ is absent in the Śivasūtras and in the root list (dhātus), but it appears in later Vedic texts, where it is accepted in sandhi rules yet never occurs as a constituent element of words. This distribution suggests that ळ emerged after the fixation of the core phonemic inventory documented in the Śivasūtras and the root system as documented in the dhātus and pratyayas — i.e., the Sanskrit language. ळ is absent in most human languages worldwide. Unlike the absence of आ (which, combined with its widespread presence and productive use in later texts, constrains the Śivasūtras to predate the articulatory capacity for a distinct long vowel in anatomically modern Homo sapiens), the absence of ळ does not impose a similar lower bound. It is included here because future research on the historical phonology of ळ may contribute to establishing more precise timelines for linguistic developments subsequent to the Śivasūtras and Sanskrit language.

Consonants

While Boë et al. (2017) demonstrate that the vocalic proto-system (vowels) is an ancient vertebrate inheritance traceable to ~25 Ma through acoustic analysis of vowel-like segments (VLSs), Demolin and Delvaux (2006) identify a distinct biological boundary for consonants. In their study of Bonobos (Pan paniscus), they conclude that no true consonants (or vowels) are observed, as these primates lack the sophisticated nervous control and the dynamic, non-uniform vocal tract shaping required to create the strictures (closures) defining consonantal production. Boë et al. (2017) do not dispute this finding for consonants and only present their finding with respect to Vowel like segments. Terminus post quem for the consonants section of the Śivasūtras is constrained at Homo-Pan split at 5–7 mya.
Consequently, the Śivasūtras preserve a phylogenetic record where the vowel section corresponds to an ancient, open-tract capability shared with Pan, while the consonant section marks the emergence of the obstructed-tract capabilities unique to the Hominini lineage post-divergence (5–7 mya).

Summing up

While the date of composition of Śivasūtras may or may not be constrained, the origination of the lstratum that Śivasūtras encode is constrained between the lower bound of 40 kya and the upper bound of 7 mya.

This conclusion is based on:

1. Presence of ऋ and ऌ as non-contact (a-spṛśya) svaras (a capacity lost in Homo sapiens)
2. Absence of आ (a capacity gained in Homo sapiens)
3. Absence of 18 varieties of svara (a capacity gained in Homo sapiens)
4. Ability to distinguish two forms of ह (a capacity lost in Homo sapiens)
5. Consonants are present in Hominina but not in Panina

Conclusion

The stratum reflected in the Śiva­sūtra cannot be younger than ~40 kya nor older than ~7 mya. This stratum is identified as Sanskrit in Pāṇini’s Aṣṭādhyāyī. Accordingly, Sanskrit—understood as this phonological stratum—must be at least ~40 kya old and no older than ~7 mya.

Under the hypothesis presented here, the Śivasūtras—the foundational building block of the Sanskrit language—constitute one of the oldest known preserved linguistic artefacts.

We believe the gap between the upper and lower bounds can be further reduced with additional work.

References

• Benton, M. J., & Donoghue, P. C. J. (2007). Paleontological evidence to date the tree of life. Molecular Biology and Evolution, 24(1), 26–53. https://doi.org/10.1093/molbev/msl150
• Boë, L.-J., et al. (2017). Evidence of a vocalic proto-system in the baboon (Papio papio) suggests pre-hominin speech precursors. PLOS ONE, 12(1), e0169321. https://doi.org/10.1371/journal.pone.0169321
• Crane, J. (1975). Fiddler crabs of the world: Ocypodidae: Genus Uca. Princeton University Press.
• de Boer, B. (2012). Loss of air sacs improved hominin speech abilities. Journal of Human Evolution, 62(1), 1–6. https://doi.org/10.1016/j.jhevol.2011.07.007
• Girard-Buttoz, C., et al. (2025). Versatile use of chimpanzee call combinations promotes meaning expansion. Science Advances, 11(3), eadq2879. https://doi.org/10.1126/sciadv.adq2879
• Hakin, B., et al. (2025). Hominins on Sulawesi during the Early Pleistocene. Nature, 625(7994), 248–253. https://doi.org/10.1038/s41586-024-09348-6
• Lieberman, P. (2012). Vocal tract anatomy and the origin of speech. In The Oxford handbook of language evolution (pp. 208–220). Oxford University Press.
• Lieberman, P., et al. (1969). Vocal tract limitations on the vowel repertoires of rhesus monkey and other nonhuman primates. Science, 164(3884), 1185–1187.
• Manser, M. B. (2001). The acoustic structure of suricates’ alarm calls varies with predator type and the level of response urgency. Proceedings of the Royal Society B: Biological Sciences, 268(1483), 2315–2324. https://doi.org/10.1098/rspb.2001.1773
• Morton, E. S. (1977). On the occurrence and significance of motivation-structural rules in some bird and mammal sounds. The American Naturalist, 111(981), 855–869. https://doi.org/10.1086/283219
• Ouattara, K., et al. (2009). Campbell’s monkeys concatenate vocalizations into context-specific call sequences. Proceedings of the National Academy of Sciences, 106(51), 22026–22031. https://doi.org/10.1073/pnas.0908118106
• Pāṇini. (n.d.). Aṣṭādhyāyī. (Traditional text; sūtra 8.4.68 referenced).
• Pāṇinīya Śikṣā. (n.d.). Traditional phonetic treatise (sūtras 4, 17–18 referenced).
• Payne, R. S., & McVay, S. (1971). Songs of humpback whales. Science, 173(3997), 585–597. https://doi.org/10.1126/science.173.3997.585
• Stevens, K. N. (1989). On the quantal nature of speech. Journal of Phonetics, 17(1–2), 3–45. https://doi.org/10.1016/S0095-4470(19)31520-7