Some Scope Creep: Difference between revisions

From The Toaq Wiki
m (reword the serialization operator idea)
(split off Weird VPs)
Line 1: Line 1:
[[Hoemaı]]'s response to [[No Scope Creep]] was a suggestion to consider serial verbs separately from other cases.
[[Hoemaı]]'s response to [[No Scope Creep]] was a suggestion to consider serial verbs separately from other cases. '''Some Scope Creep''' is a variant of that proposal which does exactly that:


Under '''Some Scope Creep''', {{t|bu kuq jí sá}} means ¬(∃x {{t|kuq jí x}}) and {{t|ao kuq jí sá}} means AO(∃x {{t|kuq jí x}}), just like in NSC — but {{t|shoe kuq}} still acts like an atomic verb as in vanilla Toaq, and so {{t|shoe kuq jí súq sá}} means ∃x {{t|shoe kuq súq x}}.
* {{t|bu kuq jí sá}} means ¬({{t|∃x kuq jí x}}), and
* {{t|ao kuq jí sá}} means AO({{t|∃x kuq jí x}}), just like in No Scope Creep — but
* {{t|shoe kuq}} acts like an atomic verb, as in vanilla Toaq Delta, and so
* {{t|shoe kuq jí súq sá}} means: {{t|∃x shoe kuq jí súq x}}.
** (In No Scope Creep, it means {{t|shoe jí ∃x kuq súq x}}.


[[File:SerialAsConstituent.png|thumb|frame|"{{t|shoe kuq}}" as a constituent. It's clear why things in TP and above make scopes but {{t|shoe}} does not.]]
So serials remain their old-school scope behavior, where {{t|shoe kuq}} behaves the same as a would-be compound {{t|kuqshoe}}.


In fact, [[Laqme]] thinks we should analyze serial verbs as constituents, rather than using movement. In that case, it's natural that {{t|shoe}} would not introduce scope; rather, {{t|shoe kuq}} can somehow be considered a V, reflecting that serials and atomic verbs ''really are'' syntactically the same. (This would change the behavior of {{t|ru}}, but perhaps for the better!)
<small>(But then how do we explain "why" {{t|shoe}} does not scope over its complement ({{t|kuq … súq sá}})? Laqme thinks maybe we really should just treat serials like {{t|shoe kuq}} as ''syntactically'' the same sort of thing as atomic verbs, and model serialization using type-shifting instead of movement. See [[Weird VPs]].)</small>
 
(I guess there is some "internal structure" to a serial V. I imagine {{t|shoe}} can "apply to" {{t|kuq}} to get the meaning of {{t|shoe kuq}}, if you do a silly amount of type-shifting. Or maybe there is a covert operator in any serial that knows how to "do" the serialization — {{t|shoe kuq}} is really {{t|shoe x<sub>shoe</sub> kuq}} where {{t|x<sub>shoe</sub>}} is some unpronounced operator that implements [[Serial_verb#Merging_definitions|merge-into]] for {{t|shoe}}'s [[frame]].)
 
So basically: scope acts as described in [[No Scope Creep]], serial verbs are just verbs, and there is no movement in VPs. This is a big step away from Chomsky-style generative grammar, towards something more like Montague grammar / [[effects|effectful]] parsing / continuations and type-checking-based semantics.

Revision as of 19:02, 25 February 2025

Hoemaı's response to No Scope Creep was a suggestion to consider serial verbs separately from other cases. Some Scope Creep is a variant of that proposal which does exactly that:

  • bu kuq jí sá means ¬(∃x kuq jí x), and
  • ao kuq jí sá means AO(∃x kuq jí x), just like in No Scope Creep — but
  • shoe kuq acts like an atomic verb, as in vanilla Toaq Delta, and so
  • shoe kuq jí súq sá means: ∃x shoe kuq jí súq x.
    • (In No Scope Creep, it means shoe jí ∃x kuq súq x.

So serials remain their old-school scope behavior, where shoe kuq behaves the same as a would-be compound kuqshoe.

(But then how do we explain "why" shoe does not scope over its complement (kuq … súq sá)? Laqme thinks maybe we really should just treat serials like shoe kuq as syntactically the same sort of thing as atomic verbs, and model serialization using type-shifting instead of movement. See Weird VPs.)