Proto-Gabrielt language

Jump to navigation Jump to search

Proto-Gabrielt (often abbreviated as PG or P-G) is the theorized common ancestor of the Gabrielt languages. Its proposed features and reconstructions have been greatly influenced by the current grammars and features exhibited by the modern Gabrielt languages, as well as historic evidence of migrations and settlements of the Gabrielt region. There is no attested record of the language ever existing, and is purely a reconstruction made by linguists beginning in the early 20th century.

A great deal of work has been carried out in the reconstruction of the language. The reconstruction and research of the Proto-Gabrielt language mainly has been done by an over-arching organization funded by the Gabrielt government called "The Linguistic Reconstruction Comittee of Gabrielland." Starting from initial reconstructions done by previous linguists in the early 20th century, modern linguists have corrected and refined the features exhibited by the early linguists in accordance to the promulgation and spread of newer theories and laws affecting the language. Many reconstruction techniques such as the comparative method were applied in determining reconstructions and more importantly sound laws which affected the different branches of the language.

It has been thought that the Proto-Gabrielt language was not a static langauge, and is in fact a period in the language's development before the attested records. Linguists now classify four stages of the language's devlopment, namely Pre-Gabrielt (3500-2000 BC), old Proto-Gabrielt (2000-1500 BC), middle Proto-Gabrielt (1500-1000 BC) and new Proto-Gabrielt (1000-500 BC). It is also theorized that the language remained a single spoken language until what is now commonly known as the "Great Split", in which the modern branches of the Gabrielt language family began to take place, although some varieties of the modern Gabrielt languages may have diverged earlier, and are the main source of insight regarding to the different stages of the proto language's development. It has also been theorized that the original speakers of the Proto-Gabrielt language lived on the southern plains of the Granya mountain range, expanding outwards at a slow pace with the advent of farming and other domesticated animals.

The migration of the original Proto-Gabrielts caused communities to split up and become isolated from one another, causing gradual shifts in the structure and overall features of the language. Once dialects, these proto-languages eventually became the ancestors of the modern regional Gabrielt languages still spoken in Gabrielland alongside the Standard Gabrielt language as the lingua franca. In each of the branches, the mutations and how the shifts happened occured in different ways, resulting in different structures and features of differing branches. Today, the Gabrielt languages are the dominant language in Gabrielland and is one of the nine official languages of the Dokodo Union.

Contrary to popular belief, Literary Gabrielt and Proto-Gabrielt are completely different concepts, with Proto-Gabrielt predating Literary Gabrielt by several hundred centuries from the first attested records of Literary Gabrielt through carbon dating. Although the name implies that Literary Gabrielt is the ancestor of all modern Gabrielt language, this is not the case as Literary Gabrielt is the mere ancestor of one single branch of the Gabrielt language which happens to dominate the linguistic scene of modern Gabrielland as the other branches of Proto-Gabrielt have become extinct or have been sidelined. Most regional languages in modern Gabrielland and the Standard Gabrielt language is a direct descendant of Literary Gabrielt.

Promulgation of the hypothesis

There is no attestation nor record of Proto-Gabrielt ever existing. Linguists who notice the similarities of features and sound shifts in the early 20th century began to create several theories about a once unified language of the Gabrielts which became fragmanted, akin to the legendary Gabrielt epics which implied that the entire Gabrielt realm "was able to understand each other as brothers in perfect symmetry". The growng trend of reconstructions around Anteria and the new methods of reconstruction popularized elsewhere increased the popularity of the single language theory. Detailed analysis of the specific changes from the proto-language into its descendant languages have not yet been fully understood, with only similar sounding words being the rationale for lexical similarity for a once united language.

This effort was however rendered futile as most academics at the time believed that the Literary Gabrielt language which, according to the legendary epic of Giliyas, "has always existed since the beginning of time", and coupled with the complex morphology and derivation Literary Gabrielt possesses, was the original and unified Gabrielt language, and no other language could be older than the ones already preserved by the respective authorities. At this period, no concrete divisions of the Gabrielt languages have been made or even suggested.

In 1908, noted Gabrielt linguist Sadi V. Karyaman listed 1,381 words across the Gabrielt languages which "sound alike", and grouped them into several branches of the "unified Gabrielt language" based on how similar a set of words sound when compared to another language. He then compared the branches to Literary Gabrielt, which has always been preserved by monks and clergymen, and found regular sound shifts in different branches which occured in slightly different ways. Karyaman noted the existence of 7 major branches of the original Gabrielt language, and proposed several reconstructions based on the sound changes. However, one branch could not be directly derived from Literary Gabrielt. In his notes, Karyaman professed that he could not create accurate reconstructions of most words with upmost certainty as one branch, particulary the northern branch, could not be described as a sound shift from Literary Gabrielt and sounded radically different and "off" from the language. In accordance with the trends of that period, Karyaman proposed that Literary Gabrielt is not "Proto-Gabrielt" as has been believed by most for centuries, and proposed an even older language predating Literary Gabrielt in which Literary Gabrielt was merely a branch of the larger family.

Karyaman's ideas recieved very negative reviews and criticism from his fellow peers, and was eventually banished from the capital by his own former friends for "insulting the gods" and "creating unnessecary commotion" with his ideas. With him being banished, Karyaman, along with an acquaintance Haruza Kasagya, set to document and record the various Gabrielt languages and comparing them with one another and with Literary Gabrielt in order to produce consistent laws regarding the sound shifts which occured in each respective branch. In 1912, after four years of field research and analysis, Karyaman and Kasagya published their sound laws of Literary Gabrielt into the modern branches of its descendants. They also published a separate set of sound laws based on the northern branches and attempted to create the first laws and precedents regarding the shifts made from Proto-Gabrielt into the branches of Gabrielt. Most of what Karyaman and Kasagya have published became actual sound laws. In 1958, in honor of the Karyaman and Kasagya, the Karyaman-Kasagya laws was named after the two.

After the publication of the 1912 sound shifts and patterns, the field of Gabrielt linguistics started to bloom. Foreign linguists, especially from neighboring countries, took interest in the works of Karyaman and Kasagya, and conducted their own research into the area. Prawasian linguist Budi Sarjono arrived in Gabrielland in 1914 and published his own theories and hypothesis regarding the Proto-Gabrielt language, supporting Karyaman's original theory. His findings got published in an international journal along with Karyaman's original work, attracting and popularizing the theory of the existence of a Proto-Gabrielt language.

With more and more linguists, both from Gabrielland and from abroad, publishing their finds about the guaranteed existence of a Proto-Gabrielt language, the original academics at the Royal Instititue of the Sciences in Vailhims, which had previously banished Karyaman from the capital, agreed to reinduct Karyaman into the society to publish his findings. More and more academics began to agree with Karyaman as well as the numerous other linguists which had been published in the field. By 1920, the precedent that a language predating Literary Gabrielt had been commonplace with the academic community, with some academics initially against the idea of a Proto-Gabrielt language publishing their own finds regarding sound shifts and the correspondences between the branches previously assumed by earlier researchers. While the vast majority of the research done conflicted with one another in respect to the specific branches the Proto-Gabrielt language has or the specific languages which fall in set branches, all studies agree on the existence of a Proto-Gabrielt language, as evidence "is irrefutible and real."

In the 1950s, after the creation of the modern Gabrielt state, reconstruction efforts were ramped up, and it is from this period that the modern classification of the Gabrielt languages were created and still widely accepted.

Historical setting

Scholars and linguists have hypothesized that the Proto-Gabrielt language was once singular language spoken around the southern plains of the Granya mountains in present-day Central Gabrielland. The earliest evidence of farming and settlements in the region came from the region, and migratory patterns could be traced using carbon-dating of objects and artefacts found surrounding the region, which consistently shows a younger and more-recent creation date than those found in Central Gabrielland. The original Gabrielts were often correlated with the Aruya culture evident in their elaborate pot-making dated from around the same time period and occuring where the original Gabrielts were hypothesized to have begin with.

With this hypothesis, it has been thought that the original Gabrielts were semi-nomads in continuous search for new lands once farming had been made impossible due to the degradation of soil quality. The constant moving in search of a better land to settle in might have caused the mass migration of the Gabrielts into newer lands. With the domestication of several animals such as the horse and the ox, the Gabrielts were able to spread out much further, increasing the distance a community had with one another, however still fairly remaining intact.

Other hypotheses include the Rielo-Guran culture hypothesis, alien intrusion hypothesis, the northern expansion hypothesis, and the Galistic culture hypothesis.

Branches

Below are the branches of the Gabrielt language family that have been accepted by the general academic community. Marginal languages with minimal attestation or controversial sources have been separated.

Branch Proto-language Description Historical languages Modern descendants
Central Gabrielt (Angyarak) Proto-Central-Gabrielt (Proto-Angyarak) Included many languages, but only descendants of Literary Gabrielt exist. Literary Gabrielt, Old Velitian, Masovian Velitian, Gasavyan, Itebean, Uyi, Etudzi, Standard Gabrielt
Eastern Gabrielt (Kitannyar) Proto-Eastern-Gabrielt (Proto-Kitannyar) Branched into three sub-families: Kitannyar, Tovadzan, and Harigyani Old Kesturikian, Hatinigyan, Old Harigyani Kesturikian, Dzaliki, Dzakvati, Harigyani
Northern Gabrielt (Tazbakan) Proto-Northern-Gabrielt (Proto-Tazbakan) Branched into Western Tazbakan and Eastern Tazbakan Old Herogyan Herogyan, Nadzuli
Southern Gabrielt (Gornyan) Proto-Southern-Gabrielt (Proto-Gornyan) Classical Gornyan Western Gornyan, Eastern Gornyan
Tezitian Proto-Tezitian All extinct, best attested language is the Hill Gari language dated to thet 10th century. Hill Gari, Harian, Beztinian None
Arama Proto-Arama All extinct Hedziya None

Marginally attested languages

The languages of the New Frontier, esepcially those in such small numbers, have been said to be a direct descendant of Proto-Gabrielt such as the Trugdzon language, however attestations are non-existent, and claims of attestations are dubious.

The original Vailhims language, originally believed to be a language which existed surrounding the capital region in the 3rd century AD, if existed, would put the language in the central branch of the Gabrielt language family. However, records of attestations are dubious or very incomplete.

Evolution

The evolution of Proto-Gabrielt from pre-Gabrielt into what is now the modern descendants of the once unified Proto-Gabrielt language consists of several major occasions both in the field of linguistics and archaelogy in respect to the migrations of the original speakers of the language. In the field of linguistics, sound changes shape the language's evolution, while in other aspects the migration and the isolation of select groups contributed greatly towards the diversion of the proto-language supported by evidence of relics and objects found and dated in specific regions of the original urheimat.

Phylogenical approach

The original groupings of the Proto-Gabrielt language, for almost all of modern linguistics, have remained vague and unclear. Preliminary models of language trees depicting genealogical relationships between the daughter languages have been made solely on a few factors, especially in the 1930s and 1940s as an early attempt to categorize the languages of the Gabrielt people. The borrowing of lexical terms (aka loan words) into several languages greatly skewed the end result of most studies, and were unaccounted in modern analysis and reconstruction of the phylogenical tree.

Through the phylogenical model, linguists predict that the Proto-Gabrielt language encountered several periods of great migration and volatility in terms of the features possessed by the langauge and the lexicon retained in its corpus. The transition from pre-Gabrielt to old Proto-Gabrielt could be marked by the splitting off of the distantly related languages of the New Frontier, however such claims are still mostly fueled by dubious claims and records of attestations, and is therefore only generally accepted due to the analogy made with other proto-languages reconstructed surrounding Gabrielland, in which substrata from neighboring proto-languages have seemingly entered the language in various forms and features.

The presence of some of these attributes and features is generally accepted to be the backbone of the modern Gabrielt language tree, in which possession of some or all of the features noted to be of alien origin equates to the language splitting off earlier in the development and evolution of Proto-Gabrielt. Branching could then be made and the relationships presented in diagram could then be conducted.

It is generally accepted that pre-Gabrielt arose sometime around 3500 BC during the first recorded migrations of the Gabrielts into the region, which then continued to evolve and marked by various sound shifts and migration patterns mapped throughout the region based on the artefacts left behind by the cultures.

Karyaman-Kasagya's law which describes the changes made from Literary Gabrielt into the descendant languages is the first such law which attempts to describe sound shifts in the Gabrielt language family. XXXX's law, based on Karyaman-Kasagya's law, attempts to describe several changes by comparing the sound shifts in Karyaman-Kasagya's law to the cognates found in languages not descendant from Literary Gabrielt.

Stages in the development of Proto-Gabrielt

The changes prescribed by linguists are a mere assumption of what the reconstructed proto-language might have undertaken during its long period of evolution and constant change. The stages prescribed in most official papers and documents document the language's evolution from its pre-Gabrielt roots up until roughly 500 BC during which a process called the "Great Split" occured, fracturing the once mutually intelligible dialects into several distinct languages, which then became their own separate branches of families with the original distinct languages serving as each branch's respective proto language.

It must also be noted that Proto-Gabrielt is not a single language, rather a period in the language's evolution which gave birth to the modern Gabrielt languages. Some languages and consequently entire branches of the Gabrielt language family have split off before the "Great Split", and is evident by the retainment of features not found in other branches which split after 500 BC.

The changes listed below occur in chronological order according to the period in which it was prescribed.

Pre Gabrielt into old Proto-Gabrielt

This stage began upon the presence of foreign influence entering the pre-Gabrielt language, with most scholars agreeing that a large quantity of the languages responsible for the shift being native languages already present before the arrival of the Gabrielts. Several notable laws and sound shifts occured during this stage. It contained many innovations, most likely through direct contact with speakers of neighboring languages, and is the first stage in the development of the language.

Merging of /r1/ and /r2/ (de-rhoticization)
  • *tʰámgʷèr1 "poison" > *tʰámgʷèr > FILLER > FILLER
  • *k1r2ùma "bark" > *k1rùma > FILLER > FILLER
  • r1èr2 "rushing waters" > *rer > FILLER > FILLER
  • The actual sounds of r1 and r2 are still a hotly debated issue. r1 and r2 reconstructions stem from the presence of two different rhotic sounds evident in two modern langauges and four attested extinct languages of the Tazbakan branch not found in any other language.
Pitch levelling of equivalent vowels (/e/ vs /i/ and /o/ vs /u/) in disyllabic or trisyllabic roots and down-accented vowels
  • *śʷéntì "oak tree" > *eśʷenti > *eśuwęti > *FILLER
  • *g2òh1r1ù "weaving needle" > *g2oh1ru > *gʲōru > *FILLER
  • *tr2áh2ʷèmkì "axe" > *tráh2ʷemki > *trāwęŋi
  • *k1ùstì "hammer" > *k1usti > FILLER > FILLER
  • /a/ whether pitched or unpitched does not become affected and is treated as a "neutral vowel".
Garonyi's law: aspirated unvoiced stop consonants become voiced and loose aspiration when fulfilling an onset role only when coda is a nasal
  • *pʰám "yolk" > *bám > *bą > *FILLER
  • *k1ʰéni "house" > *g1ʰeni > *gʰęį > *FILLER
Normalization of laryngeals (h1 and h2)
  • Word initial laryngeals when preceding a consonant which isn't a nasal are lost — *h1lir1a "soul > *lira
  • Laryngeals preceding pitched /a/ are lost
    • *h1ástʰim "bake" > *ásdim
    • *h1àr2èn "sugar" > *aren
  • Laryngeals preceding /e/ and /i/ in nucleus position are lost except when coda is a nasal
    • *h2r1ésʷi "crocodile > *résʷi > *resuwi
    • *h1ika "window" > *ika
  • In reconstructons of the masculine accusative ending -Vh > -Vi
    • *pʰám-ah "yolk" ACC. > bám-ai
  • Laryngeals are still present when serving as the intervocalic link between two vowels
Epenthesis of /ś/ when preceding a stop and linking to another stop suffix
  • The sequence evolves as /FT/ > /FśT/ > /śś/
  • *h2lep-tʰau "folk" > *lepśtʰo > *leśśo
Final accented vowels lose their accent and an epenthic /s/ is added
  • *g2ʷené "moon" > *g2ʷenes > *guwenes
Epenthesis of /e/ when following a fricative — *śʷéntì "oak tree" > *eśʷenti > *eśuwęti
Tariman's Law: Geminate consonants in affixation are reduced
  • coronal with coronal /TT/ > /s/ — *h1mit-dù "father-in-law" > *misus
  • dorsal with dorsal /KK/ > /h1/ — k1lég2-ka "forest" > k1léh1a
  • labial with labial /BB/ > /m/ — ap-bʰé "route" > ames
Unstressed -au becomes /o/
  • *h2lep-tʰau "folk" > *lepśtʰo

Several major changes to its grammatical morphology and features have been extrapolated from the remaining samples of the more conservative and ancient Tazbakan languages of northern Gabrielland when compared to the rather simple grammatical morphology of languages that split after the 500 BC. For one, the simplification of the noun class system in pre-Gabrielt into old Proto-Gabrielt could be seen in the reduction of declensions and amount of endings a noun could possess depending on its gender and the declension case it falls into. Once having 7 genders, the masculine I, II, and III and feminine I, II, III were all merged into a singular masculine and feminine class respectively, eliminating 4 classes, leaving behind only vestiges of the split masculine and feminine system in old Proto-Gabrielt, whose remnants have already well deteriorated into middle Proto-Gabrielt.

A similar phenomena could also be seen in the verb system of pre-Gabrielt into old Proto-Gabrielt. The optative and subjunctive moods once both present are in the process of merging in old Proto-Gabrielt, and have completely merged upon the new Proto-Gabrielt era. The mediopassive verb, however, was still retained, albeit getting generalized along the process. While there were more than 10 ways to form the infinitive in pre-Gabrielt, only two survived into old Proto-Gabrielt. By the time of old Proto-Gabrielt, the tense and aspect system had been split into its now familiar components.

Old Proto-Gabrielt into middle Proto-Gabrielt

This stage began when the accented vowels once prominent in pre-Gabrielt began to be watered down in favor of a simpler stress system. Middle Proto-Gabrielt is also theorized to have birthed long vowels and nasal vowels into the mix, giving the language a total of 15 vowels when only viewing quality and not pitch as nasal vowels are not able to become lengthened. The pitch system would be comlpetely lost in the transition into new Proto-Gabrielt. The consonant system began to mutate and shrink, with the loss of most laryngeals in the language as well as the tremendous shifts in the stop consonants as described in Marjorie's law. Reduction also played part in the shifting of the sounds, resulting in several major innovations.

Merging of h1 and h2 into a singular laryngeal sound
  • *tʰémih1e > *demihe
Loss of palatalization and labialization
  • Palatalized consonants /ʲV/ > /ijV/ — *tʲòh2sú "to bark" > *tʲoh2su > *tijohsu
  • Labialized consonants /ʷV/ > /uwV/ — *tr2áh2ʷèmkì "axe" > *tráh2ʷemki > *trāwęŋi
Lenghtening of vowels when preceding a laryngeal
  • Vowels preceding laryngeals in /Vh/ or /VhC/ get lengthened — *tr2áh2ʷèmkì "axe" > *tráh2ʷemki > *trāwęŋi
/z/ > /s/ before dental consonants
  • EXAMPLE
Marjorie's Law: Major shift in the realization of stops. Unvoiced aspirated stops affected by Garonyi's law still applies.
  • Voiceless unaspirated plosives become as follows:
    • /k1/ > /ś/ — EXAMPLE
    • /k2/ > /x/ — EXAMPLE
    • /p/ > /w/ — EXAMPLE
    • /t/ > /y/ — EXAMPLE
    • When preceded by a sibilant, only /k1/ and /k2/ follow Marjorie's law
  • Voiceless aspirated plosives become as follows:
    • /k1ʰ/ > /k/
    • /k2ʰ/ > /c/ — EXAMPLE
    • /pʰ/ > /p/ — EXAMPLE
    • /tʰ/ > /t/ — EXAMPLE
  • Voiced unaspirated plosive become as follows:
    • /g1/ > /ʑ/ — EXAMPLE
    • /g2/ > /ɣ/ — EXAMPLE
    • /b/ > /pʰ/ —
    • /d/ > /tʰ/ —
  • Voiced aspirated plosives become as follows:
    • /g1ʰ/ > /g/ — EXAMPLE
    • /g2ʰ/ > /ɟ/ — EXAMPLE
    • /bʰ/ > /b/ — EXAMPLE
    • /dʰ/ > /d/ — EXAMPLE
Hush's Law: Fricative interaction in various scenarios
  • Fricatives are voiced when preceding a pitched syllable, causing the syllable to be unpitched and become stressed
    • /s/ > /z/ — EXAMPLE
    • /ś/ > /ʑ/ — EXAMPLE
    • /x/ > /ɣ/ — EXAMPLE
  • Voiced fricatives at the coda of a stressed or pitched syllable are unvoiced, causing the syllable to become unpitched
    • /z/ > /s/ — EXAMPLE
    • /ʑ/ > /ś/ — EXAMPLE
    • /ɣ/ > /x/ — EXAMPLE
Nasalization of vowels in various circumstances
  • /Vm/ > /Ṽm/ > /Ṽ/ — EXAMPLE
  • /Vn/ > /Ṽn/ > /Ṽ/ — EXAMPLE
  • /Vnk/ > /Ṽŋ/ — *tr2áh2ʷèmkì "axe" > *tráh2ʷemki > *trāwęŋi
  • /Vng/ > /Ṽŋk/ — EXAMPLE
  • /VnC/ > /ṼC/ — EXAMPLE
  • Intervocalic nasals nasalize the previous vowel but are not omitted to serve as the intervocalic link
Unstressed o becomes a (/o/ > /a/) — EXAMPLE
Unstressed /au/ and /ou/ merge into /ō/ — EXAMPLE
Unstressed /e/ becomes /i/ — EXAMPLE
  • /e/ when part of a diphthong is preserved unless the diphthong in question is /ei/, in which /ei/ > /ī/ — EXAMPLE
Word final m merges with n in stressed syllables (/m/ > /n/) — EXAMPLE

A radical change in the way verbs are presented is thought to have happened in the middle Proto-Gabrielt era. Verbs began to gain aspectual markings for more circumstances when compared to old Proto-Gabrielt in the form of complex affixes attached to verb. The set of affixes present in middle Gabrielt have been attempted to be recreated by comparing the present affixes in modern Gabrielt languages, however the set of affixes present in middle Proto-Gabrielt is thought to have arisen from accompanying words, and recreating them has been more challenging than mere recreations of the affixes. The affixes in question now are active productive contributors towards minor sound shifts in the language and is perhaps an even greater contributor in the new Proto-Gabrielt language.

The optative and subjunctive grammatical moods have fully merged in middle Proto-Gabrielt. Remnants of different affixes attached to different verbs to indicate a subtle change in meaning reflect the once productive optative mood. The subjunctive mood, however, is perhaps more dominant in speech than before as vestiges of the subjunctive mood in Literary Gabrielt could be seen in the form of a rich and varied conjugation system for the subjunctive, a feature absent from most languages. It could be said that the subjunctive is just as fleshed as the indicative in middle Proto-Gabrielt as it has been in Literary Gabrielt. These claims are the most widely-accepted theory.

The inanimate and neuter genders are also showing a process of merging. It has been thought that the animate and neuter genders become merging upon the discovery of more monotheistic religion, which suggests that the Gabrielts then thought, with the advent of monotheism, that the world around them is inherently inanimate, and subsequent difference in treatment from the neuter is redundant. By new proto-Gabrielt, it has been thought that the two genders have completely merged with only the most frequently used words becoming an irregularity within the declension paradigms of the neuter gender.

Middle Proto-Gabrielt into new Proto-Gabrielt

New Proto Gabrielt is probably the phase in which the most reconstruction work has been done in due to the abundance of attestations in various langauge near the theorized "Great Split" sometime around the 6th century BC. During the transition into new Gabrielt, the radical changes of middle Proto-Gabrielt are kept and underwent even more shifts, with some shifts defining the general "feel" of the modern Gabrielt languages occuring in the transition between middle and new Proto-Gabrielt. Pitch has completely been supplanted with stress, and the appearance of more long and nasal vowels dominate the scene. Many unstressed variants and phonemes become merged into specific sounds.

Mutation of geminate consonants
  • fricatives /SS/ > /tS/ —
  • stops /BB/ > /sB/ —
  • nasals /VMM/ > /ṼM/ —
  • liquids /LL/ > /jL/ —
Unstressed /Vwa/ becomes /Vu/
Mutation of o into u in penultimate stressed syllables (/o/ > /u/) —
Loss of aspiration by compesatory lengthening
  • The loss of aspiration via compensatory lengthening is by far the largest conrtibutor towards the once absent long vowels in proto-Gabrielt.
Loss of word-final voiced stops
Epenthesis of /t/ preceding /s/ in stressed syllables
/w/ > /v/ within onsets —
Some open syllables in word-final positions become nasalized
  • This phnomenon is especially stark in the area of case endings and affixation, where the majority of the change is first observed.
Loss of /j/ in between vowels