deaı

[Ntsekees]

  {
    "toaq": "deaı",
    "type": "predicate",
    "english": "▯ is worth/deserving of satisfying property ▯.",
    "gloss": "worth",
    "short": "",
    "keywords": [
      "worth",
      "deserve",
      "deserving"
    ],
    "frame": "c 1",
    "distribution": "d d",
    "notes": [],
    "examples": [
      {
        "toaq": "Dẻaı ní lî kảqgāı da.",
        "english": "This is worth seeing."
      },
      {
        "toaq": "déaı",
        "english": "those who are deserving of something"
      }
    ],
    "fields": [
      [
        "raı"
      ],
      [
        "cuaq"
      ]
    ]
  },

What exactly is the meaning of that word?

Currently it looks like a hybrid of the two following meanings, which I think should be kept separate:
• satisfying property ▯ is worthy / would be beneficial for ▯. (currently juqgīshē in Toadua);
• ▯ deserves/merits/earns the rewards of satisfying property ▯ for the service of satisfying property ▯ (basically Lojban's {jerna}; currently juqdēaı in Toadua).

—Ilmen.

[uakci]

The examples point towards the first interpretation, in which case the raising is suspicious. (There's nothing wrong with ___ deserves / is worthy of being the case.) And the second interpretation would be useless to have if we already have deaı.

[xorxes]

It sounds like Esperanto -inda, which is worth having. The unfortunate part is that it is most useful with second-slot properties (worth-having, worth-reading, worth-eating, worth-thinking-about, worth-sharing, worth-doing, etc,) so for serial predicates it will usually be deai mu ...

[Ntsekees]

@xorxes: In that case, deaı would presumably take a binary relation argument, in the fashion of kıu.

[xorxes]

-ebla and -enda are Esperanto's companion "passive" suffixes, so yes, like kiu and duai.

English has its -ebla: doable, understandable, visible, audible, measurable, etc.

and an -enda of sorts: a must-do, a must-have, a must-buy, a must-try

As usual, I don't really understand this binary relation you want. All the slots of the second predicate remain as slots of the resulting predicate except that the one selected by the lambda variable changes its meaning.

kiu fuba X Y Z ... -> kiu X li fuba ja do Y Z ...
duai fuba X Y Z ... -> duai X li fuba ja do Y Z ...
deai fuba X Y Z ... -> deai X li fuba ja do Y Z ...

It's either an ordinary 1-slot property, or it takes an n-ary relation of the same arity as fuba. I don't get why it should be binary.

[Ntsekees]

  {
    "toaq": "kıu",
    "type": "predicate",
    "english": "▯ is ▯able.",
    "gloss": "-able",
    "short": "",
    "keywords": [],
    "frame": "c 2",
    "distribution": "d d",
    "notes": [],
    "examples": [],
    "fields": []
  },

kıu2 is a relation; there are two reasons for that:

1. it avoids the need for mu by accessing the second slot of the argument predicate along with its first slot;
2. it prefills the first slot of the argument predicate with an impersonal value, presumably "baq pỏq".

kỉu chủq ní = relation_able(this, λx,y: eat(x,y)) = this is eat-able (edible). (the eater slot is abstracted away and not explicitly filled, it presumably corresponds to the impersonal pronoun "one" or impersonal "you", i.e. "one/you can eat this").

dảı mủ chủq ní sa (rảı) = dảı ní lî chủq sa (rảı) ja (rảı) = it is possible that something eats this.

If I'm not mistaken, dảı mủ chủq ní is currently not allowed because mủ chủq is necessarily a binary predicate (mu forces a binary reading and not an unary reading for chuq) but then only one argument (ní) is provided to the whole serial; the value of the second argument cannot be left elliptically implicit as it may be in Lojban.

[xorxes]

Would that mean that if I wanted to use kiu as the predicate by itself I would have to place two ja in the second slot, and kiu would silently bind the first one to an implicit há and the second one to its first slot? Or does that only apply when used in serial predicates?

kíu ní chûq ja ja instead of kíu ní chûq há ja?

[xorxes]

The question here seems to be whether the semantics of deai, duai, kiu force a ha argument into the property/relation that goes into its second slot, or whether it leaves that slot open to be filled by something else.

deai (mu) moi "__ is worth thinking about" or "__ is worth thinking about by/for __ "
duai (mu) moi "__ has to be thought about" or "__ has to be thought about by/for __ "
kiu (mu) moi "__ is thinkable" or "__ is thinkable by/for __"

Also, is "mu" needed there, or do those predicates shuffle the slots inside their second argument by themselves?

Are there other predicates like these that tie their first argument to what is typically the second slot of their property argument instead of to the first?

I guess actually duai can go both ways:

duai moi "__ has to think" or "__ has to think about __ "
duai mu moi "__ has to be thought about" or "__ has to be thought about by/for __ "

kiu has deq for the agent side, Is there a corresponding predicate for duai like deq is to kiu?

[Ntsekees]

With kıu2, the agent slot is the first one, so with moı there's no need for mu: kỉu mỏı hóq = "that is thinkable".

I think there's no incompatibility between the impersonal kıu/2 version (binary) and the kıu/3 version (ternary) allowing specifying who is able: arity polymorphism allows distinguishing them according to the number of arguments provided to the predicate.

kıu is the impersonal equivalent of deq (or possibly daı, I'm not quite certain).

duaı is patterned after deq, and not kıu.

I don't know of an existing kıu-patterned equivalent of duaı, but with a serial predicate you could use fıhā dủaı gủjēo, "one/you must be, with ▯, in relation ▯".

I guess actually duai can go both ways:

duai moi "__ has to think" or "__ has to think about __ "
duai mu moi "__ has to be thought about" or "__ has to be thought about by/for __ "

I don't think the second one makes sense if what has to be thought about is inanimate, and it is probably often unsuitable when it is animate. The first participant is the one which has the duty/obligation to satisfy the property; an inanimate object cannot be morally bound to fulfill an obligation.
With dủaı mủ mỏı súq jí, it is you who is under the obligation to make me think about you, it's not me.

[xorxes]

I think we're not understanding each other, I never meant to suggest the existence of a kiu/3.

kiu has two arguments. We are in agreement about that. The question is what does it do when it acts as the left component of a serial predicate. Does the resulting combined predicate have one fewer argument than the second predicate of the serial, or does it have the same number of arguments as the second predicate of the serial?

You interpret the effect of kiu in serials as reducing the number of arguments of the second component by one (it eliminates the agent slot).

moi has two slots, therefore you say kiu moi must have one slot.

teq has four slots, therefore kiu teq presumably must have three: "X is payable to Y for Z"

I was questioning whether it was necessary/useful for kiu to remove the agent slot.

Assuming kiu must do that for some reason, another question is whether deai must have the same behavior, and another question is whether duai is also like that.

I think the biggest problem in this discussion is that you are coming from a frame of mind of "things are like this (about kiu) and there's nothing to discuss about that", and I'm still at the "what would be the best/most convenient way for kiu to behave?" stage.

For you it is unthinkable that kiu's frame could be anything but c 2 while I am arguing that it might perhaps be better if it was c 1.

[xorxes]

As for duai, it can be impersonal in the same way that kiu can be impersonal. Whatever possible worlds kiu existentially quantifies over, there's a corresponding version with the universal quantifier.

If something can be thinkable, possible to think about, it could also be "think-enda", not possible to not think about,

kiu noaq "... is readable from ... (by ...?)"
??? nuaq "... is to be read from ... (by ...?)"

Etymologically "legend" has this kind of suffix in it.

[Ntsekees]

As for kıu, I was merely explaining its official behavior as I understand it. kıu was apparently meant to parallel the suffix -able in behavior, by removing the agent slot. Although in English the agent slot can be optionally reintroduced with the preposition "for": "it's unthinkable for you".
An alternative to kıu which would achieve a similar effect with binary predicates in serials, allowing expressing the agent in the second slot, would be mu dẻq (and not deq mủ), although it doesn't have the impersonal reading of kıu and would give awkward argument orders when applied to ternary predicates. If we are to keep the agent impersonal, then kıu mỏı can be rephrased as fıhā dẻq mỏı, "one can think about ▯", which is also unary.
A c 1 kıu-like could be fıhā dẻq tủa jẻo ~ jeotūadēqhā, one is able to make ▯ satisfy property ▯: jeotūadēqhā shỉe hó = "he is awake-makable, awakable"; however with binary predicates, this will requires mu where kıu didn't require it.

As for duaı and deaı: I think duaı should remain c 1, which is the simplest form. fıhā dủaı… is an option for mimicrying the behavior of kıu in serials. With regard to deaı, I'm not certain what exact meaning it is meant to have, so I can't tell which type signature it should be having.

Incidentally, I think duaı is not about ineluctablility/inavoidability, but about the fact that bad social consequences for the duty-holder if the duty is not carried out.

[xorxes]

I found one word that doesn't work well with the official behavior of kiu because it's arguments are backwards for some reason: cheaq

kiu cheaq would not mean "trustable", which would have to be kiu mu cheaq.

deai cheaq would work for "trustworthy" if deai behaves normaly, but almost every other predicate would require a mu.

So what's the dual of huai, which would mean something like socially permitted/acceptable? These words about possibiity/ability/capability/permissibility on the one hand and necessity/ineluctability/compulsivity/obligation on the other should always come in pairs, I think.

[Ntsekees]

There's dı.