This time, we’re getting into the morphology. Well, kind of – no affixes will be made today. However, I want to get a rough sketch of the fundamental ideas behind this language. In other words, I’ll be making charts, but won’t be filling them out yet.
Let’s start with perhaps the most important feature of protolang 1: verb agreement. Like Yiksighe, I want protolang 1 to be highly head-marking, and specifically, I want it to feature polypersonal agreement using a direct-inverse system. I won’t go into the full details of what that entails, but here’s the basic idea:
- Core arguments are always ordered based on animacy, relevance, or some other ordering, regardless of which one is agent and which is patient.
- Proto-lang 1 uses this common order: 2nd person > 1st person > 3rd person proximate > 3rd person obviate > 3rd person inanimate (the so-called “animacy hierarchy”)
- If the agent comes after the patient, a special inverse marker is applied to the verb to tell you that.
In Yiksighe, I had both the focus (the higher-animacy argument) and the non-focus (the lower-animacy argument) marked as prefixes. In protolang 1, however, I want the non-focus to be marked as a suffix on the verb. This will lead to some prefixes and suffixes taking alternative forms, as I discussed in the last two posts.
You’ll notice that I mentioned an obviative 3rd person above. Obviation cooccurs very commonly with direct-inverse systems. I’ll use a 3-part 3rd person distinction, with proximate animates, obviative animates, and inanimates (which aren’t marked for obviation). Inanimates also will be unmarked for plurality. This yields the following table:
| |
Singular |
Plural |
|
1st
|
|
|
|
2nd
|
|
|
| 3rd proximate |
|
|
| 3rd obviative |
|
|
| 3rd inanimate |
|
In addition, there are some forms that act like person markers, but won’t have the function of person markers. The most important of these is the indefinite marker. In the protolanguage, transitive verbs always need to take two affixes. This can be a problem for constructions where only one argument is important – for example, “I’m eating”, or “It is being eaten.” In this case, one uses a special indefinite person marker. While it carries the basic meaning of “someone/somthing”, it can be used as a passive, antipassive, or even impersonal. It’s lower on the animacy hierarchy than any of the personal affixes.
The reflexive and reciprocal voices also function like personal affixes – replacing the non-focus argument. You can’t use the inverse with a reflexive or reciprocal, though, just like you can’t say “myself sees me” in English.
Another feature I’m keeping from Yiksighe is the relativiser prefix (or, as I called it in Yiksighe, the “attributive person.”) This basically is a relative pronoun attached in place of the focus, making the verb usable in relative clauses. It’s also how adjectives are formed in this language. There’s no distinction in proto-lang 1 between adjectives and verbs (e.g. the word for “red” actually means “to be red”). The relativiser prefix can thus be used to form proper adjectives out of adjective-y stative verbs.
The relative pronoun that got agglutinated inflected for number and animacy, and it kept this inflection when it was affixed. That means that relativised verbs, and therefore adjectives, agree with the noun they’re modifying, a pretty unusual trait in verb-like adjectives.
Prepositions in protolang 1 are also verb-like. They inflect in the same way as intransitive verbs. They can take the reflexive and reciprocal affixes (as in the english “to themselves” or “to each other”), or the relative affix (as in the english “to whose”).
Although this isn’t agreement, per se, I think it’s important to talk about the noun system. The nouns are much more simple than the verbs: animate nouns are marked both for obviation and plurality, inanimate nouns are marked for neither. (However, when it must be specified that they’re singular, they take a singulative particle to clarify.)
The only other marking nouns take is the construct state. As protolang 1 is so head-marking, I thought that it would be fun to use the construct state, as I rarely see it in conlangs. If you aren’t familiar, a construct state is kind of like a reverse genitive – a special marking on a possessed noun to specify that it’s possessed. If the possessor is already known (or is in the 1st or 2nd person), a personal prefix can be applied to the noun to indicate it. For the sake of simplicity (and not having to come up with more affixes), I’ll say that these prefixes are the same as the focus prefixes for verbs. The construct state also overlaps interestingly with obviation. As possessees are often obviative, we can say that the construct state automatically makes a noun obviative. This means that construct state marking and obviation marking can’t both be present on the noun. (Inanimate nouns can still take the construct state, but it doesn’t make them obviative – they’re still inanimate.)
Obviously, this is a very rough starting point. We haven’t even considered tense, aspect or mood, let alone some of the other features I want to add to this language. But I hope that this should give a feel of my goals for protolanguage 1, and for the language family as a whole.