17. Ꝡé fı faq hóa / What's about to happen
Conversation
Íqse: Chum geı súq sá shatı naına ꝡeı.
Rúa: Obe, kıjı! Luı jea jí ní hao tî sá gıaqtue.
Íqse: A, ına, hıq gaı jí ké fuaq, ꝡë pam hóa shátı.
Rúa: Pu fı ruaq máo jí, ꝡá esagı, ꝡé chum geı súq hóa.
Íqse: Zı, zı. Tam jea jí sá fuq, ꝡë bu kuo hóa ba…
Rúa: Súq, jü chuabo hóa ké suqbo, nä bo hóa sá rıq ba!
Vocabulary
| Word | Meaning |
|---|---|
| chum verb | verb-ing |
| tam verb | perfective aspect |
| luı verb | has verb-ed |
| hıq verb | has just verb-ed |
| za verb | has yet to verb |
| fı verb | is just about to verb |
| shatı | ▯ is a shirt |
| sueta | ▯ is a sweater |
| geı | ▯ wears ▯ |
| Word | Meaning |
|---|---|
| naına | ▯ is cool/stylish |
| gıaqtue | ▯ is a concert |
| fuaq | ▯ is a picture |
| pam | ▯ is (e.g. printed) on ▯ |
| esagı | ▯ rules / is awesome |
| cem | ▯ is normal |
| tıao | ▯ suits ▯ |
| chuabo | ▯ is called ▯ |
Aspects
In grammar, aspect tells us how the time an event happened relates to the time we’re actually referencing.
For example, “I will be sleeping” and “I will have slept” both refer to the future. But in the first sentence, the sleeping event surrounds that future time, and in the second sentence, the sleeping is before that future time.
The default aspect is tam, which views the event as a whole and places it within the reference time. Even though it’s already implied by default, you can say tam explicitly to really emphasize this “point-event” perspective.
Pu tam faq sá luaı.
Something funny happened.
The very common aspect word chum expresses the progressive aspect, like the English “-ing” form. This means the reference time is contained within the event time.
Pu chum koı íme.
We were walking.
Aspect semantics
The “reference time” is the time interval t contributed by the tense particle, and an event has a temporal trace τ(e) telling us the interval it plays out over.
Then, for example, in jıa luı nuo jí:
- nuo jí part says there was some sleeping-event e
- jıa says that the reference time t is in the future,
- luı bridges the gap between the two, saying τ(e) < t.
Inertia worlds
Consider the sentence: “I was logging in, when suddenly the computer crashed.” We can’t really analyze this as referring to a complete logging-in event, as no such event exists. We are refering to what would have been a complete logging-in event in an alternative world where the computer didn’t crash.
Such alternative worlds — where the things that stop us from what we’re doing don’t happen, and the future is maximally compatible with the past — are called inertia worlds. They were introduced by David Dowty in his 1979 book Word Meaning and Montague Grammar to explain sentences like the above.
In Toaq, pu chum haqbaı jí is analyzed like so:
For all inertia worlds w branching off from the past time t, there is some cooking-event in w such that t ⊆ τ(e).
This way, we can use the progressive aspect even for events that never quite fully take form in the world as we’re discussing it.
More aspects
The aspect luı expresses that the event “has” happened, and now the reference time is in its aftermath. That is, the event caused some state change that is still true at the time referred to.
Pu luı gom jí.
I had made a mistake (and now I was suffering the consequences).
The aspect hıq is like luı, but the event has just finished.
Pu hıq chuq jí sá karı.
I had just eaten some curry.
The aspect za means the event is yet to happen at the time referred to.
Pu za geq nháo jí.
She had yet to meet me.
The aspect fı means that, at the time referred to, the event is just about to happen.
Pu fı rıufa nháo ké doaq.
She was about to return to the city.
Relative clauses with ꝡë
The particle ꝡë goes after a noun, and starts a relative clause that restricts it. Inside the clause, you can refer to that noun using hóa.
Sıgı tú kue, ꝡë luı fıeq nháo hóa.
Every book that he has written is interesting.
There isn’t an English word for this hóa, because in English, relative clauses just have a “gap” in them where the hóa would go.
Ma zao súq ké poq, ꝡë marao jí gâq hóa tî ké patı?
Do you know the person that I danced with at the party?
You can also use other pronouns bound by the noun the clause is attached to.
Cho jí ké fuaq, ꝡë pam{hóamáqfúaq}ké shatı.
I like the picture that's on the shirt.
In the rising tone, ꝡé is short for ké ꝡë (the thing which…)
Cho jí, ꝡé pam hóa ké shatı.
I like what’s on the shirt.
Restriction and non-restriction
By “restrict”, I mean that ꝡë whittles down the domain that the article selects from. This sentence makes a generic claim ranging only over chocolate from Mexico, not just any chocolate:
He cho jí báq choko, ꝡë sıao Méhıkogua hóa.
I like chocolate which comes from Mexico.
There is another relative clause particle, jü, which starts a non-restrictive relative clause. The following sentence says that I like chocolate, and incidentally, chocolate comes from Mexico.
He cho jí báq choko, jü sıao Méhıkogua hóa.
I like chocolate, which comes from Mexico.
This subtle difference is marked by the mere presence or absence of a comma in English! In Toaq, you always write a comma, but use a different particle to start the clause.
Relative clauses on pronouns
The word jü is usually what you want when attaching a relative clause to a name or pronoun. It just doesn’t make sense to “restrict” such an argument, because it already refers to exactly one person.
Fı nuo Láqme, jü seakuaı hóa.
Laqme, who is tired, is about to sleep.
Donkey sentences
A donkey sentence is a sentence that contains a variable outside of its scope island. It’s named after the following example:
(?!) Tú poq,ꝡë bo póq sá aqshe,nä haqdo póq áqshe.
Every person who owns a donkey feeds it.
These sentences are hard to account for when translating Toaq into logic formulas, because such occurences are disallowed in formal logic.
∀[X: poq(X) ∧ ( ∃[Y: aqshe(Y)] bo(X, Y) )] kıaı(X, Y)
These sentences are a real thorn in semantics’s side. They are clearly grammatical in English, but it’s very difficult to explain why.
For a while, Toaqists wondered if speaking a loglang means learning not to produce donkey sentences. Should we just say: “every person feeds each donkey they own”? But it turns out that these “donkey pronouns” are really everywhere in everyday speech. Can’t we just make them work somehow?
These sentences more or less spawned the entire framework of discourse representation theory, which does away with the tree-recursive transform of clauses into first-order logic sentences, and explains this it as referring to a binding in a mental “discourse structure” that updates as the listener reads or listens their way through a sentence.
Giving up so much of the classical theory is a shame. There are some theories of natural language syntax that try to explain the donkey anaphora phenomenon “as-is”, and there are in turn some theories of Toaq syntax that try to adopt those explanations. One of these is based on continuations: see Donkey anaphora is in-scope binding by Barker and Shan (2008).