Proto-??

Consonants
Proto-?? had a three-way contrast between plosives, differentiating between voicless, voiced and aspirated stops, and differentiating between labial, alveolar and dorsal stops. There were also labialised varieties of the dorsal plosives. Despite the large variety of plosives, the other areas of Proto-?? consonant morphology were lacking, with there being only two fricatives, two semivowels and four sonorants in the language. The sonorants m, n, r and l could appear in syllabilic form.

Vowels
Proto-?? had only two vowels and two semivowels, but distinguished between long, short and medium-length vowels. Vowels formed an important part of Proto-?? morphology, with all vowels playing a role in the ablaut paradigm, used for creating participle forms of some roots and for pluralisation of verb and noun forms.

Pronouns
As Proto-?? was a pro-drop language, distinguishing between all persons and all numbers in verbs, pronouns were typically used only for emphasis in the nominative, but normally for other cases. Proto-?? did not have third-person pronouns as such, but rather had a large body of demonstratives which could also be used to convey the sense of he, she, it and they.

Roots
The most important part of any Proto-?? morpheme was its root. Roots conveyed the basic and broadest meaning of any morpheme, without denoting whether the morpheme was a noun or a verb, and without conveying any grammatical information. Many different nouns and verbs could be formed from the same base root, through reduplication, ablaut or suffixation.

Participles
Roots were generally converted into participles to convey a perfective sense (in verbs) or to form nouns and adjectives. Participles were generally formed from their roots by reduplication, ablaut or suffixation, with ablaut sometimes being employed simultaneously with one of the other two categories. The participle suffixes were -s, -t, -k, -p, -v, -u̯ and -i̯, and suffixation was considerable more common as a means of participle-formation than the other categories. There were two ablaut paradigms, e-grade ablaut and o-grade ablaut, with the grades named after the middle vowels in both paradigms. The paradigms were i̯ > ě > e > ē > ō and u̯ > ǒ > o > ō > e, and typically affected the first vowel of the root. Reduplication, the final method of participle formation, was the most variable, as it could manifest itself as minimal reduplication (reduplication of only a single phoneme), partial reduplication (reduplication of a single syllable) or full reduplication (reduplication of the whole root). The reduplicated section could also be inserted at most positions in the root. It was occasionally accompanied by dissimilation, producing participles that appeared very different from their roots.

Verbal
Proto-?? verbs were generally formed through simple addition of endings to the base root, though some verbs did take inflected stems as their base. These different endings denoted tense, mood, grammatical person and number. Proto-?? verbs belonged to two declensions, denoted by their infinitives, -i̯voH and -esoH, and every verb declined for seven tenses (with the present, aorist, imperfect and future tenses being formed from simple inflection of the unmarked verb stem and the perfect, pluperfect and future perfect tenses being formed from addition of endings to verb participles), four moods (indicative, imperative, subjunctive and optative), three persons and three numbers (singular, dual and plural).

The present expressed action happening in the present time, as well as in the immediate future. The aorist was used to describe actions happening in the past with no continuing relevance. The imperfect denoted action initiated in the past and continuing into the present, action initiated in the present and continuing into the future, reoccuring past action or action initiated and terminated in the past but with continuing relevance in the present. The future was used to describe events destined to take place after the time of speaking. The present perfect, pluperfect and future perfects were formed through the addition of present, imperfect and future tense endings to the participle stem. The subjunctive was used to express counterfactuals and hypotheticals, and was formed through the addition of the subjunctive/optative marker -ṛ(i̯) to the verb stem and through ablaut. The optative mood, used to express desire and wishes, took the same ending, but did not undergo ablaut. The imperative, used to from commands and occasionally questions, took different inflections from the indicative, and did not decline for the first and third-person singulars. The person endings were broadly the same throughout the indicative, subjunctive and optative paradigms, with -ti̯ (except in the subjunctive, when it was simply a vowel), -di̯ and -si̯ representing the first, second and third persons respectively. The addition of an n before the consonant (forming the clusters -nt, -nd and -ns respectively) was used as a form of pluralisation. Verbs were atypical in that ablaut was used, not in order to form the plural (though it was used to form the dual), but in order to form different tenses.

Nounal
Proto-?? nouns were formed through the combination of a root and an ending, with more complex construction being formed through the combination of a stem and an ending. This latter group represented a more common formation than in verbs, and many different nouns could be formed from the same base root in this way. Clitics and reduplication could be used to form new nouns (oskoH > siskoH, with initial vowel loss accompanying the reduplication). Some suffixes used to form nouns from roots were ṃ, ṇ, ṛ, ḷ, tṃ, tṛ, Vd, Vt and Vk.

Proto-?? nouns declined for eight cases, which were nominative (used to denote the subject of a verb or the object of a predicate), accusative (used to denote the object of a verb or with certain prepositions), genitive (used to denote a relationship between nouns, such as for possession, in the genitive of the whole or for forming adjectives), dative (used to denote indirect objects), locative (used to denote that something was at or near something else, with certain prepositions or for representing the time at which something took place), lative (used to denote that something was moving towards something else, with certain prepositions denoting movement towards somewhere else and for representing the time until something took place), ablative (used to denote motion away from something, with certain prepositions relating to this function and for representing the time from which something took place) and the instrumental (used to demonstrate the means by which something was achieved). Proto-?? nouns also declined for three genders, masculine, feminine and neuter, with gender being an innate quality of every noun. Only a few nouns (generally occupations, people and animals) could take multiple genders. In addition to this, there were two separate noun declensions, marked by their different nominative endings, -oH/-o/-e and -oi̯o/-ei̯o. A root could potentially form two separate nouns by using each declension (peroH and peroi̯o from the same root word per-, producing noun couplets in daughter languages such as pero and piroia). There was no distinction between the plural and dual neuter endings.