Phonology of Sanskrit

Sound: The fabric of our reality?

Auṃ ॐ

As per the Hindu cosmogony, this primordial sound is what the supreme deity Brahmā who emerged from a lotus from the navel of lord Viṣṇu (विष्णु) uttered which generated the five great elements Pañcamahābhūta (पञ्चमहाभूत) :

1. Ākāśa (आकाश): Ether/Space first element formed from the vibration of Aum.
2. Vāyu (वायु): Air from ether's motion
3. Agni (अग्नि): Fire from air's friction
4. Jala (जल): Water when fire condenses
5. Pṛthvī (पृथ्वी): Earth when water solidifies

The universe emerges from nāda (नाद) the primordial sound. Nāda Brahmā (नाद ब्रह्मा) i.e. the sound of creation is considered the foundation of existence.

Bhartṛhari (भर्तृहरि) was a Hindu linguistic, philosopher, and a poet circa 5^th century CE. He is best know for his work Vākyapadīya (वाक्यपदीय) which is treatise on sentences and words. In the first verse of his book he states:

अनादिनिधनं ब्रह्म शब्दतत्त्वं यदक्षरम् ।

विवर्ततेऽर्थभावेन प्रक्रिया जगतो यतः ॥ १ ॥

anādinidhanaṃ brahma śabdatattvaṃ yadakṣaram |

vivartate'rthabhāvena prakriyā jagato yataḥ || 1 ||

"The Brahman who is without beginning or end, whose very essence is the Word, who is indestructible, who appears as the objects, from whom the creation of the world proceeds."

Across different cultures we see a similar phenomena. Christianity places a big emphasis on the power of sound through words.

"In the beginning was the Word, and the Word was with God, and the Word was God. Through Him all things were made; without Him, nothing was made that has been made."

In Christian worships practices, believers connect to the divine through hymns, psalms, and chants. Gregorian chant was foundational to early Western music influencing the development of polyphony, musical notation, sacred compositional traditions, etc.

The act of creation in Judaism too is linked to speech. Jewish tradition teaches that the world was created through ten divine utterances. The utterances from Genesis include "Let there be light" and other statements from God. The blowing of shofar a trumpet made from the horn of a kosher animal with the marrow removed during Rosh Hashanah (the Jewish New Year) and at the conclusion of Yom Kippur (Day of Atonement) shows the significant role played by sound in Jewish rituals.

In Islam, the Quran describes the act of creation as. being initiated by the divine command Kun which means "Be". The full phrase is "Kun Fayakun" meaning "be, and it is". Appearing 8 times in the Quran, it symbolises the instantaneous nature of divine creation: when Allah wills something to exist, it exists immediately.

In Sikhism, Shabad (Divine Word) and Naam (Divine Name) are central to the creation of this world. Hukum (Divine Command/Will/Order) governs the creation. Guru Granth Sahib the holy scripture of Sikhism talks about the practice of Kirtan, the singing of hymns as a a form of devotion to foster a deep connection with the Creator.

Several Native American cultures believe that songs and sounds are not just tools for communication but are sacred and shape the world. The Hopi people identifies the role of singing in their cosmology where the Creator Taiowa first conceived the world and Sotuknang his nephew executed the creation through the power of sound and song. The Lakota people have ceremonies like the Sun Dance or the use of the Sacred Pipe with scared songs that are believed to sustain the cosmic order. The songs are believed to be "language of the spirits", allowing for a direct connection with the divine. For the Navajo people the Holy Wind (Níłch’i) is the source of all movement, breath and speech. Chanting in ceremonies is designed to re-align the individual with the Holy Wind to restore beauty.

In many African cultures drumming sounds, chanting and singing plays a big role in spiritual practice. The Dogon people of Mali of West Africa believe that Creator Amma, began the universe with vibrations and is often compared to the weaving process where sound and vibration are the threads that create the fabric of reality. These vibrations are symbolised by drums which is a sacred instrument that plays the rhythms of creation. The Yoruba people use Oriki, praise poetry to call forth the spiritual power of the ancestors into their physical realm. It is believed to carry the power of creation as it invokes the energy of the Orishas (deities). Across many African cultures, the drum is a scared instrument that bridges humanity to divine. the Djembe Drum is believed to carry the heartbeat of earth. The drumbeat is associated with the pulse of life.

In Chinese philosophy, Daoism (Taoism) and the concept of Dao (Tao) are tied to sound and vibration which is the ultimate principle that governs the universe. Sound and vibration are seen as manifestations of the Dao's flow and harmony. Japanese Shinto and Buddhist traditions too reflect the belief in the power of vibration and resonance. In Shinto Kotodama "the spirit of words" empahsises that spoken words carry spiritual power.

Ancient Egyptians and Greek had similar outlook of sound being the force in the creation of this universe. As per the Egyptian Memphis creation myth, the creation is attributed to Ptah which was not a physical but an intellectual creation by the Word and the Mind of the God. Ptah's creative thought and speech were believed to have caused the formation of Atum, the primordial God from whom all arose and Ennead a group of 9 deities including Atum. The Greeks believed that the universe was governed by an intrinsic order often described as Kosmos which was reflected in music, mathematics, and the natural world. Spoken word held an utmost importance in the Greek mythology. As seen by the myths of Orpheus, the legendary musician and poet who could charm animals and even influence gods with this powerful songs and words.

I hope you get where I am going with these examples and analogies with respect to sound and our quest to make sense of our reality and the creation of this universe. Humans are phenotypic creatures, we focus on the manifested form in the sense of what we see - more than what we hear. Audition is secondary while sight is primary. Shifting our attention to sound reveals it as the primary medium through which we have historically deciphered reality. To speak is not merely to label, but to instantiate - to execute a universe through the power of utterance. Language therefore, is not just a tool for communication it is the very architecture of the human experience.

Introduction: Phonology, the study of sound patterns in language

Language is a structured system of communication consisting of grammar and vocabulary. Language forms the primary means by which humans convey meaning in spoken, signed, and written forms. There are various theories around how language evolved but the truth is we do not know; all theories are just speculation. But that doesn't stop us from not studying this phenomena and trying to understand its intricacies. The study of language is called linguistics. But since language is such a complex phenomena, linguistic is not confined to study just one aspect of language. The areas of linguistic analysis are syntax (rules governing the structure of sentences), semantics (meaning), morphology (structure of words), phonetics (production of speech sounds), phonology (the topic of this section), and pragmatics (how the context of use contributes to meaning).

Language begins with sound, when most of us think of language we think of speaking. If there is no sound there is no language (exception being sign language). Phonology is the study of the categorical organisation of speech sounds in languages. It studies how sounds are organised in the mind and used to convey meaning. Then there's phonetics not to be confused with phonology. Phonology studies the abstract ideas of sound and how we use those sounds to communicate whereas phonetics is concerned with the physical act and properties of sound. We make variety of sounds using lips, tongue, mouth and some bits of our throat. If you want to make let's say a [b] sound, you close your lips first and then open it and release the air all at once. Phonetics deals with how we make this sound, how we hear it, how we can describe it deal with them empirically, and how we can make it visible to its practical to write it down and print.

Phonology investigates how sounds are organised in a particular language and how these sounds follow rules and a predictable pattern. These investigations are also focused on how the sounds are stored and processed in the brain and what's their function in conveying meaning in the linguistic contexts. Phonology forms an interesting part of social reality as phonological patterns - how individuals use sound patterns - can indicate a lot about them, examples being their social class, regional background, educational background, ethnicity, personal identity, etc. A subfield of phonology, called sociophonetics intersects sociolinguistics explores how variations in phonological patterns correlate with social factors.

Phonology also affects language acquisition. Babies start to tune to their native languages quite early and are quite proficient at recognising and producing phonological patterns which is crucial for the development of grammar and vocabulary. Phonology from a cognitive perspective explains the mental processes involved in speech production and perception. The words stored in the brain are not just organised by semantics but also by phonological structures and this organisation affects how we retrieve and process words. Native speakers of tonal languages like Mandarin, Cantonese, and Vietnamese are more likely to have a precise and stable form of absolute pitch perception and reproduction.

Studying phonology offers a window into our minds, language and society. It shows that sound systems are not merely abstract entities but deeply connected to our social identities, cognitive processes, and our inquiries into the human mind.

Phonemes

Phoneme is the fundamental unit in phonology. A phoneme is a smallest unit of sound that can change the meaning of a word. In English for example, the sounds /b/ and /m/ are distinct phonemes because they distinguish meanings of the words like "bat" and "mat". Phonemes are abstract unit of sounds and represent a group of related sounds that speakers of a language perceive as identical. There are slight variations in how those sounds in a group are pronounced based on context and these variations are called allophones. For example the sound [p] in words "pot" and "spot". When you say pot you leave some air out of your mouth (called aspiration) after making the [p] sound but when saying spot there's no aspiration.

Note: You'll see me using two notations to describe phonemes. The notation /p/ is the phoneme as it exists in the brain as a category. The notation [p] is the phone the actual physical sound coming out of our mouths. Even though they are same phonemes /p/ there are two different phones [p] (without an aspiration) and [p^h] (with an aspiration). Do not pronounce these sounds as alphabets like Pee or Bee but more like Pa or Ba.

Now that we have covered some basics of phonology, let's understand what are the sounds we need to make in order to speak a language like Sanskrit.

Indo-European: The mother tongue of Sanskrit

We all carry the past around with us all the time. We have inherited our physical features - eyes, face, lips, hair, and nose - alongside deeper biological predispositions from food sensitivities and allergies to our neurological response to fear and anxiety. In genetics, these are our heritable traits; we are the living manifestation of our ancestors' biological history. It's a physical blueprint passed down from people we've never met. I hardly know anything about my great-grandparents let alone their names. If you think about it, it's quite unsettling that we do not much about our ancestors from the past who have given us so many of our traits.

Archaeology is our best hope and way to acknowledge the humanity and importance of the people who lived before us. It is our best effort to investigate the past lives not described in writing which constitutes majority of human history. Humans have spent 95% of their time as hunter gatherers, we are biologically and phonologically calibrated for a world we no longer live in. We have absolutely no clue of how our hunter gatherer ancestors spoke and communicated. Sounds do not fossilise, we cannot know the specific vocabulary or grammar of hunter gatherers from 50,000 or 100,000 years ago. However, that does not stop us from searching for the language that our recent ancestors spoke. Many linguists believe that language we use today is the medium to know the lives of our ancient speakers. It contains fossils of of their lives and environment which we can decode. If we can reover their vocabulary through the words we use today we can get an insight into the everyday lives of the communities that lived in the past. In fact, a substantial vocabulary list has been reconstructed for one of the languages spoken about five thousand years ago and is the ancestor of modern languages like English, French, Spanish, German, Hindi, Marathi, Punjabi, Nepali, Bengali, Persian, Italian, Russian, Polish, Latvian, Irish, Welsh, Scottish Gaelic, etc. to classical languages like Latin, Avestan, and Sanskrit, ancient Greek to ancient languages like Anatolian, Tocharian. All the languages shown in the figure below descended from this same mother tongue belong to one family: Indo European language family.

Today these languages are spoken by more than 3 billion people. The origin of people who spoke this mother tongue has been a topic of fierce debate since about two hundred years now and has evolved into a cultural and racial war ever since the idea was hypothesised. From Eurocentrics to Indocentrics, nationalists and dictators have attempted to claim their country as the homeland of the people who spoke this mother tongue. So, how was this mother tongue idea materialised?

In 1771, Sir William Jones a Welsh philologist published his book Grammar of the Persian Language and was the first English guide to understanding the Persians. This book at the age of 25 earned him a reputation as one of the most respected linguists in Europe. His translations of Persian poems inspired many European Romantic philosophers. He was appointed as one of the three justices of rhe first Supreme Court of Bengal in Calcutta. He was to regulate both the English merchants and the rights and duties of Indians as the colonial subjects in the British Raj. The issue he faced soon after arriving was that although the English merchants recognised his legal authority, the Indians obeyed an already functioning and ancient system of Hindu law which was regularly cited in the court by the Hindu legal scholars called pandits. The English judges had no idea if these laws actually existed since most of the legal texts were in Sanskrit a language foreign to the judges. Jones started to read the texts and made comparisons not just with Persian and English but also Latin and Greek which he had learnt in university, with Gothic a literary form of German which he had also learnt and with Welsh his native tongue. On February 2 1786 he made the following announcement in the Asiatic Society of Bengal which he had founded when he had arrived to India:

"The Sanscrit language, whatever be its antiquity, is of a wonderful structure; more perfect than the Greek, more copious than the Latin, and more exquisitely refined than either, yet bearing to both of them a stronger affinity, both in the roots of verbs and the forms of grammar, than could possibly have been produced by accident; so strong indeed, that no philologer could examine them all three, without believing them to have spring from some common source, which, perhaps, no longer exists."

While ancient civilisations have studied grammar, the 19^th century marked the birth of Comparative Philology which was the precursor to the modern linguistics. When the idea of Indo-European as a language group was introduced we only knew how to define the language family and how to determine which languages belonged to the family. The discipline of linguistics and the early analytical methods invented by the 19^th century philologists are still used today to describe, classify, and explain language variations across the world. Historical linguists provided us with this ability to reconstruct parts of extinct languages for which no written evidence survives today. This is possible by relying on the regularities in the way sounds change in our mouths. Let's say if you collect the Indo-European word for hundred from the different branches of the language family and jot down the pattern of sound change in all of those words you can reconstruct a single hypothetical ancestral word as the root for that word in all of the branches. The Latin word for hundred kentum and the Sanskrit word śatam are cognates of the ancestral root word ^*ḱm̥tóm- (The asterisk * is a "orthographic signal" in linguistics. It tells the reader: "This word is never found in writing; it is a mathematical reconstruction"). This is a hypothetical root for the word hundred in Proto-Indo-European (PIE) mother tongue. It's a statistical average rather than a spoken word. We don't know how the word was pronounced as sounds don't leave any fossils, we can't hear the past. Linguists have reconstructed the sounds of more than 1500 Proto-Indo-European roots. So how do we know that they are accurate? Well we can't but on the other hand, archaeological excavations have revealed inscriptions in Hittite, Mycenaean, Greek, and archaic German that contained words, never seen before showing the precisely the sounds previously reconstructed by comparative linguists. This sort of gives credibility to the linguists and their reconstructed roots. Linear B was a syllabic script that was used for writing in Mycenaean Greek which is the earliest attested form of the Greek language. When Linear B was deciphered in 1952, it revealed a form of Greek much older than Homer. It contained a "kw" sound that linguists had reconstructed years earlier but had never actually seen in later Greek texts.

So, yes that the reconstructed roots and words are not real but at least they can be considered a close approximation and that gives a framework to work on the comparative study of the language family. The reconstructed lexicon is a window into the environment, social life, and beliefs of the speakers of Proto-Indo-European. The lexicon reconstructions show that they had words for cattle (cow, ox), otter, beaver, wolf, lynx, elk, horse, mouse, hedgehog, goose, duck, eagle, bee, honey, sheep, pig, dog, etc. The lexical evidence can be attested and compared against archaeological remains to reconstruct the environment, economy, and ecology of the speakers. So, how do linguists reconstruct such languages that we have never heard or seen in a written form?

Phonological change: How to reconstruct a dead language

Most of the languages in the Proto-Indo-European language family can be divided into 2 categories based on how the the languages represent the number 100: Centum and Satem. Centum comes from Latin and is pronounced as Kentum. Latin C is pronounced as K and this is a bit confusing for people learning Latin languages Like French and Spanish. In fact it also confuses people learning English since English uses Latin script. Satem comes from Avestan which is the holy language of Zoroastrians. Avestan and Sanskrit are daughters languages and share a lot of similarities among them. But why create such a category based on how these daughter languages represent the number 100? The answer is the sound difference. This division is based on how the Indo-European languages treat the sound coming from the back of the throat, also called as velar which is the soft palate at the back of the mouth and include phonemes like [k], [g], etc. The satem languages are the eastern division of the language family geographically speaking and include languages like Indo-Iranian, Armenian, Albanian, and Balto-Slavic. The centum languages are in the western group and include Greek, Italic, Germanic, and Celtic. These two words are cognates as they both mean "100". However, much like any other category of taxonomy we come up with there are always exceptions. Tocharian (now extinct) an Indo-European language from modern day northwest of China in Xinjiang is part of the centum family. The word for 100 in two of the dialects of the language was känt (Tocharian A) and kante (Tocharian B). So, this division is merely academic and does not actually show a clear geographic divide within the Indo-European language family.

Before we dive into how linguists reconstruct a dead language let's clear up some terminologies and concepts used in linguistic.

Stops vs Fricatives vs Affricatives

Pay attention to the highlighted sounds in these words:

• b in bat
• s in sun
• j in judge

These 3 sounds sound so different when you say the words. These sounds are the 3 most important type of sounds in human language: Stops, Fricatives, and Affricates respectively. In phonetics, consonants are classified by how the airflow is obstructed through the vocal tract which is called the manner of articulation. Articulation is how we shape or obstruct the air flow from our lungs as it leaves our body to create different sounds. If the airflow is completely blocked and then released we get a stop which is also called a plosive. If the airflow is partially obstructed creating friction we get a fricative. If you combine the two - block first, then release into friction - we get an affricate. Some examples of each in English:

• Stops: [p], [b], [t], [d], [k], [g]
• Fricatives: [f], [v], [th] or [θ], [s], [z]
• Affricates: [tʃ] 'ch' in chat, [dʒ] 'j' in judge

The character ʃ is called Esh and it represents fricative, eg. 'sh' in ship.

The character ʒ is called Ezh and is the voiced twin of Esh, meaning your vocal chords vibrate when you say the j in judge.

The phonemic system of IE contained vowels, semivowels, and consonants. The IE consonant system consisted mostly stops.

The Latin word for 100 is Centum (pronounced Kentum) and sometime in early Medieval French there was a dialectical version of this word which was tsent'm. The sound change that has occurred here is called palatalisation. In instances where c (k) a velar sound was preceded by a front vowel - vowels produced when the tongue is positioned towards the front of the mouth like i and e - went through a softening process and turned into a dental affricate [ts]. By the time modern French evolved the [ts] sound was simplified further and turned into a sibilant [s], so the word for 100 in French is cent (pronounced with s). Latin cera [kera] 'wax' became French cire pronounced seer and Latin civitas [kivitas] 'community' became French cité pronounced seetay. These sound changes were not random and confined to certain words, they spread systematically to all similar sounds in the language. In instances where Latin k- was followed by back vowel (-o) it remained a k-. For example, Latin costa 'rib' is French côte.

Human language is governed by rules and they determine the sentence construction (syntax), the relationship between sounds of words (phonology and morphology), and their meaning. The direction of sound changes are governed by two constraints: ones that are generally applicable across most of the languages, and those specific to a single language. The general constraints are imposed by the biological and mechanical limits of the human vocal anatomy. Constraints within a language are imposed by the limited range of sounds that are acceptable and meaningful for that language. Using this piece of knowledge linguistics can determined which phonetic variants came first and which ones came later. But how do they make these decisions?

Two general rules help figure out the order. One is that initial hard consonants like k and g tend to move towards soft sounds like s and sh if they change at all but a change from s to k would be quite unusual. Another one is that a consonant pronounced as a stop in the back of the mouth k is more likely to shift toward the front of the mouth t or s in a word where it is followed by a vowel that is pronounced in the front of the mouth. This phenomena is termed as assimilation: one sound tends to assimilate to a nearby sound in the same word, simplifying the movements needed to say the word. The specific type of assimilation of k to s is called palatalisation. In fact, palatalisation has been a key factor in development of French from Latin and is responsible for much of the unique phonology of the French language.

Assimilation usually changes the quality of a sound and sometimes removes sounds by combining two sounds together. The opposite of assimilation is the addition of new sounds to a word. Many native English speakers insert [-uh] when they say athlete. Instead of two syllables Ath-Leet it turns into three: Ath-uh-Leet and introduces a schwa sound uh. The schwa sound is also known as the lazy sound.

The two phenomenon explained above are called phonological and analogical changes and are the mechanisms through which new forms are incorporated in a language. Linguists examine several different points in the past - inscriptions in say classical Lati, vulgar Latin, early Medieval French, later Medieval French and modern French - to map the all the phonological and analogical shifts in the evolution of French from Latin. Let's now understand how the word for 100 was reconstructed for Proto-Indo-European.

The first thing linguists did was that they gathered up all the daughter words of the Indo-European language family in a list. Here, knowing the rules for the sound change is of utmost importance as some sounds can change radically. Let's take hundred as a reference.

The linguists asks: are these words phonetically transformed daughters of the a single parent word? If yes, they are cognates. Also, to prove they are cognates they need to be able to reconstruct the sequence of phonemes that could have developed into all the documented daughter sounds through the known rules.

Let's look at the first word. We have 3 variations in the sounds: [k], [h], [s]. The [k] in Latin can be explained if the parent PIE term began with a [k] sound as well. But what about Avestan, Old Indic, and Lithuanian [s]. Remember hard consonant sounds tend to soften through palatalisation so if the parent PIE began with a [k] then it is quite plausible that these language groups went through a sound change and turned to a sibilant [s]. Well why can't we argue that the sound shift happened from [s] to [k] and that PIE had [s] as the first sound? Going from a sibilant fricative (s) to a velar stop (k) is rare for several physical reasons:

• Complexity: A [k] sound requires a total blockage of airflow at the back of the mouth (the velum), followed by a sudden release of air. A [s] sound is a continuous stream of air pushed through a narrow gap.
• Energy expenditure: Stops like [k] and [g] require more muscular tension in the tension and throat. We tend to naturally gravitate towards path of least resistance which is why [k] often softens into [s] or [sh].
• Articulation: Sounds usually move forward in the mouth (palatalisation). Moving a sound from front to back is unnatural for the tongue in rapid speech.

So we can add more weightage to the possibility that [k] was the first sound in the parent PIE word but what about the [h] in Gothic, Old Saxon, Old English, etc.? These daughter languages are in the category of Germanic languages, English is a Germanic language which uses Latin script.