Proto-Beltonic

Proto-Beltonic or Common Beltonic is the hypothesized common ancestor to the Beltonic languages, which include Beltonian, Wistarian and Lennodesian.

The language was spoken in southeastern Beltonia between 1000 BC and 500 AD, before spreading more widely across the region during the Beltonic Migration. Verifiable records of the language are limited to occasional inscriptions in Beltonian Runes and transcriptions by Greek and Roman writers. The primary means of reconstruction has been the comparative method.

Proto-Beltonic features nonconcatenative morphology; its roots are defined not as fixed syllables but by a fixed group of consonants that can have varying vowels between them. For example, the root *j-l-m gives rise to the words *jalám (to weave), *julám (past tense of to weave), *julúm (imperative mood of to weave) *jalím (weaver) and *juláma (woven cloth). However, the language also features words with concatenative morphology, suggesting that it was in a state of transition to the latter.

Notable differences between Proto-Beltonic and its successors are its sparse inventory of vowels and wider inventory of consonants, including some that appear to be unusual uvular or laryngeal consonants, though it lacks a contrast between voiced and voiceless stops. Proto-Beltonic was highly inflected, comparable to Old Beltonian and Classical Beltonian but more complex than any other descendants.

Consonants
Proto-Beltonic had 25-27 consonant phonemes. Only the fricatives had contrasting voiced and voiceless forms, and they contrasted only on the start of words and in inflectional suffixes. Stop consonants tended to be voiced in the middle of words, especially after /u/ and emphasized vowels.

There are two or three uvular or laryngeal consonants whose place of articulation cannot be easily identified. These are denoted as h1, h2, h3. The existence of h3 is more controversial but supported by a majority of recent scholarship on the grounds that it explains why Old Lennodese continued to maintain a glottal fricative even as some were lost. All three phonemes had allophony, producing sounds ranging from palatal to glottal fricatives; some of them were released as [ɸ] in the dialects that became Lennodese. Bacha suspects that the original forms were respectively /h/, /χ/ and /ħ/, while Smithson instead gives them as /h/, /x/ and /ɣ/.

Late Proto-Beltonic may have seen the assimilation of /j/ into consonants before it in non-emphasized syllables. These changes were [sj] becoming [ʃ], [tj] becoming [tʃ], [dj] becoming [dʒ] and possibly [zj] becoming [ʒ]. These were allophones that were likely not realised as fully post-alveolar consonants and did not merge with the /ʂ/ and /ʐ/ phonemes.

Vowels
Proto-Beltonic had three monophthongs and two diphthongs; some reconstructions count /ja/ and /wa/ as diphthongs. In addition, each vowel phoneme had both an emphasized form as a well as a standard form. It is unclear as to whether the emphasis came through a pitch accent, vowel length, simple emphasis or some other method, but the pitch accent is the most accepted theory. The emphasized and standard forms only contrasted in inflections based off the same root. In reconstructions, the emphasis is represented by an acute accent. This implies a high tone in IPA, although it may have been something else.

It is possible that /e/ and /o/ existed as allophones of /i/ and /u/ as they did in the earliest forms of Beltonic and Wistarian. However, Ansleh argues that there were too many differences in those language to suggest that these allophones had a common origin.

Phonotactics
Proto-Beltonic had few consonant clusters. On the syllable onset, the only consonant clusters permitted are those where the second sound are /j/ or /w/. Vowels are the only permitted syllable nucleus. On the syllable coda, the only consonant clusters permitted are some which are created by inflections, where the secound sound is /s/ (including its allophone [z]) as well as /nt/, /lt/ and /rt/. Ansleh argues that non-emphasized /a/ was generally omitted in late Proto-Beltonic if it was followed by a liquid consonant and another vowel. Nonetheless, at least some of these /a/ vowels survived in Old Beltonian without being immediately omitted or reduced.

Roots
The Proto-Beltonic roots came in three types: bilateral (two consonants), trilateral (three consonants) and quadrilateral (four consonants). It is possible that late Proto-Beltonian had quinqueliteral roots (five consonants) which for the most part behaved as a quadrilateral root with a prefix that was unaffected by the word's inflections. The majority of words came from trilateral roots, though some common words have bilateral roots. A root may include no actual consonant at the start or end, although some linguists like Smithson have suggested that these were glottal stops.

Inflections focus on the one or two core vowels of the root. Quadrilateral roots divide into two types. The most common are prefixed roots, in which the last two consonants are core vowels, while others are suffixed, in which the first two are core vowels. The two core vowels behave similarly to those of a trilateral root, but the non-core vowel may also be affected.

Bacha makes two theories about the origins of words: the oldest words in the language were bilateral roots, and that the original vowels of most roots are /a/, though she notes some clear exceptions such as *h2uk "pig" (which may have been onomatopoeic) and *dudur "foreigner". The original form is usually a noun, but in some cases such as *j-l-m it may be a verb. She further notes that many trilateral nouns became a verb through u-mutation to mean using the noun, or through i-mutation to mean obtaining in the noun, but this relationship was not inconsistent.

Cases
Proto-Beltonic had six or seven cases: ergative, copulative, absolutive, dative, locative, instrumental and vocative. The existence of a vocative case in Proto-Beltonic is disupted. If it did exist, as it did in early Beltonian and Wistarian, it would have mostly resembled the absolutive case. Only pronouns changed for all cases; nouns inflected for five sets of cases, while adjectives inflected for three.

Proto-Beltonic had ergative–absolutive alignment. The ergative and absolutive cases correspond to the nominative and accusative cases of languages with nominative–accusative alignment, with the exception that nouns that were the subject of an intransitive verb took the absolutive case. The copulative case is an unusual feature of the language, which corresponds to a noun and pronoun with the present tense of the copula and is also used in some tense formations. It appears to have been formed by the merger of absolutive nouns and pronouns with forms of a previous copula.

Sound mutations
Proto-Beltonic had three key sets of sound mutations, which played noticeable roles in inflections.

I-mutation
I-mutation works by colouring the first core vowel that is not /i/ nor proceeded by /j/ starts as following: In addition, if one of the core vowels is /i/, the emphasis moves to this vowel.
 * /a/ becomes /ja/
 * /u/ becomes /ju/
 * /wa/ becomes /wi/

U-mutation
The u-mutation is the mirror of the i-mutation. There is no evidence that u-mutation had a role in inflections, but it did influence the relationships between words, such as between nouns and verbs. In addition, if one of the core vowels is /u/, the emphasis moves to this vowel.
 * /a/ becomes /ja/
 * /u/ becomes /ju/
 * /wa/ becomes /wu/

Cross-mutation
The third form of mutation focuses on swapping both certain sounds and emphasis. Until i- and u-mutation, this mutation can only affect both core vowels simulatenously.
 * /i/ swaps with /u/
 * /j/ swaps with /w/
 * The emphasis moves to the other core vowel if there is one

Nouns
Nouns inflect for 3 grammatical numbers and 6 sets of cases, if the disputed vocative case is included. The conceptual form is used when the noun is used as a concept rather than to refer to a specific object or set of objects. The standard plural forms appear to have been formed by the merger of the conceptual plurals with a definite article, which is reconstructed as *tan based on other demonstrative pronouns that start with the same sound. With the i-stem weak nouns, the conceptual form was originally the same as the singular, but at the point when Lennodese diverged, its conceptual forms were switching to represent the standard plural forms.

Nouns use the same inflection for the dative and instrumental cases. It appears that early Proto-Beltonic had a dative case that merged into the instrumental case except with some pronouns. Additionally, the vocative case is the same as absolutive case, except that the plural takes the same inflections as the conceptual form.

One ongoing topic of discussion is as to whether Proto-Beltonic is descended from a language that lacked a plural form or whether the conceptual form was originally a standard plural.

The nouns are divided between strong nouns, which tend to be older roots with a nonconcatenative morphology, and weaker nouns, which tend to be newer roots with a concatenative one.

Strong nouns
Strong nouns follow an older pattern. The defining characteristics are the use of i-mutation to form the plural and the cross-mutation to form the ergative. The other cases are form by similar suffixes to the weak nouns.

Bilateral strong nouns work similar to trilateral ones, with one exception: if their sole consonant is not affected by the mutation, then a suffix is added (/i/ for i-mutation and /a/ for cross-mutation). Quadrilateral strong nouns always have the last two vowels as the core vowels.

Anomalous nouns
There are at least four nouns that appear to be suppletive, with singular and conceptual/plural forms coming from separate roots. Furthermore, there are at least six where the conceptual/plural forms are formed by irregular mutations, with three dealing with intimate family relations.

Pronouns
Pronouns used similar cases to nouns. The difference is the existence of the dative case, which has not merged into the instrumental case except for the first and third person plurals. The instrumental, locative and vocative cases produce the most regular formations, with the latter almost entirely identical to the absolutive case, except for the second person plural.

The third person plural pronouns appear to be related to the demonstrative pronouns and the defunct article *tan and demonstrate the most regular formation. Aspects of the strong noun inflections also appear to affect the feminine third person singular. Ansteh argues that the ergative forms of the other third person singular pronouns originated from their absolutive forms mutating via previous uvular approximants; there is no dispute that the remaining pronouns have separate roots for the ergative case.

Verbs
Like the modern Beltonic languages, Proto-Beltonic had multiple moods. Those that are documented are the indicative, energetic, jussive, imperative, conditional and potential moods. Others may have existed, but did not survive in the successor languages. In addition, the standard past tense (preterite) functions as a mood. Each of these forms has separate inflections for the two numbers and three genders of the subject (but not the grammatical person) as well as an infinitive form, present participle and past participle.

Compound tenses formed with the participles tended to imply that the event was more continuous, either in the present or past. These were formed using the subject in the copulative case. The preterite was generally used to describe a single event. In late Proto-Beltonic, the compound tenses took precedence and the original indicative present became a habitual mood.

It is not clear how the future tense was formed, as its immediate descendants developed different ways of forming a future tense. Smithson suspects that for the most part, the present tense was used for future events, but when disambiguation was needed, one of the auxiliary verbs *qifu ("to go, to walk") or *rah1a followed by the preposition ku (together: "to stand to, to be prepared to"). The former became the basis for the future tense in Beltonian, while the latter become the equivalent for Wistarian.

Similar to nouns, strong verbs tend to be older words which followed the older pattern and weak verbs, which tended to be later additions to the language.

Adjectives
Adjectives agree with the gender, number and case of the noun they describe. Unlike nouns and verbs, there is no evidence of any strong or irregular patterns. The plural and conceptual take the same forms, and cases are grouped into three sets. In late Proto-Beltonic, the copulative forms began to take the ergative forms.

The case endings mean that the placement of adjectives is flexible; they do not even have to be adjacent to the noun. The mergers of certain words suggest there was a preference for adjectives to follow the noun, but this was not universal.

Adverbs
Adverbs are most often similar to an adjective, but with the inflectional endings removed. Bacha writes that "on first sight, it appears to be a paradox that Proto-Beltonic adjectives are adverbs with added endings, whereas in most languages the opposite is the case." The explanation for this is that in early Proto-Beltonic or its predecessor language, adjectives and adverbs were identical, but adjectives subsequently acquired inflectional endings. In some cases, the final consonant of the adverb experienced lenition such as *zit ("quickly") and *zitsu ("fast").

Syntax
The case system of Proto-Beltonic probably caused the language to have a free word order, which persists to varying extents in its descendant languages. The formation of the copulative case suggests that the language had a default SVO word order, at least up until that point. Later, it appears that VSO word order become common in order to permit the mergers of the participles with the copulative pronouns.