User:Cchar94/sandbox

Koopman's (1989) Logical Variables

Koopman (1989) proposes that "kO" is not a complementizer, but instead is its own verb and taking a sentential complement.

Koopman proposes that n-pronouns (also used as logophoric pronouns) are logical variables and are Ā-bounded by an operator at C (Koopman 1989). A feature like [+/- n] could account for the binding distinction between the two series of pronoun, and be used to account for logophoric effects seen in Abe with n-pronouns. whereas o-pronouns are [- n] and n-pronouns are [+ n], and featural conflict would account for the binding distinction of the two series of pronouns that is summarized below (Koopman 1989, p. 567).

An O-pronoun cannot be coindexed with an n-pronoun if the C of the first clause containing n c-commands o.
(Koopman 1989: 570)

O-pronouns cannot be c-commanded by an antecedent that is an n-pronoun
N-pronouns must be co-indexed with an antecedent that is a n-pronoun, and must be disjoint from a reference that is not a n-pronoun (Koopman 1989: 562)

In other words, [+ n] elements can never be bounded by [- n], and vice versa. Thus, the operator that binds n-pronouns is also [+ n]. If there is an operator that is [+n], it could account for the distribution of the two series of pronouns in Abe, as summarized below: The [+n] operator would bind n-pronouns, and o-pronouns cannot be bound by the npronoun or by the noperator because of feature conflict.

The logophoric effect is seen when the complementizier kO is introduced as a verb with a [+n] silent subject.

The complementizer kO originates as a verb with a [+n] silent subject, which can only bind with [+n] elements like n-pronouns. [+n] elements cannot bind [-n] elements like lexical NP, QPs, and O-pronouns and vice versa. Based on Chomsky's Theta criterion, the silent subject will receive the theta role of the agent, which will manifest itself as the semantic property of logophoric pronouns binding with the speaker of an indirect discourse. Koopman proposes that if this can be generalized to other languages, then there are no "logophoric pronouns", but instead two distinct mechanisms coming together to give the effect of logophoricity.

An O-pronoun cannot be coindexed with an n-pronoun if the Comp of the first clause containing n c-commands O n-operator c-commands an O-pronoun, the O-pronoun and the operator must be contraindexed.

a purely syntactic analysis of the distribution of n-pronouns vs o-pronouns in Abe:

A verb that aselects a kO-complement assigns the feature [+n] to its designated argument. Koopman proposes that instead of as a pure complementizer, kO in Abe is introduced as a verb heading its own clause and takes a sentential phrase as its complement (Koopman 1989, p. 583). If so, the verb kO would also be introducing a silent subject in its specifier position, and receive the same theta-role the verb "say" assigns to its subject. Koopman then proposes that the silent subject selected by kO carries the feature [+n], which then liceneses the binding effects. Due to certain LF restrictions/requirements, the kO then moves to the Comp position of its CP.

Logophorcity

The notion of logophoricity describes the behaviour of certain pronouns that must bind with an antecedent "whose speech, thoughts, feelings, or general state of consciousness are reported" (Clements 1075). Claude Hagège first introduced the term logophoric pronouns in 1974 after studying many African languages that includes a separate paradigm of a distinct class of pronouns (as opposed to indirect reflexive pronouns). Logophoricity is also used to account for non-clause-bounded reflexive pronouns found in Japanese and Icelandic.

There are some syntactic analysis to account for logophoricity within the current Binding Theory, as well as semantic approaches using the Discourse Representation Theory. However, it usually involves a combined theory of the two to fully account for the distribution of logophoric pronouns.

The term logophor was introduced by Claude Hagège (1974) after observing that many African languages use a distinct class of pronouns to refer to the source of indirect discourse: the individual whose perspective is being communicated, rather than the speaker who is relaying this information. George N. Clements (1975) expanded upon this analysis, arguing that indirect reflexive pronouns serve the same function as logophoric pronouns, even though indirect reflexives do not exhibit a distinct form from their non-logophoric counterparts, whereas logophoric pronouns (as Hagège defined them) do (Reuland 2006, p. 3). For example, the Latin indirect reflexive pronoun sibi may be said to have two grammatical functions (logophoric and reflexive), but just one form (Clements 1975). More recent analyses of logophoricity are in line with this account, under which indirect reflexives are considered to be logophors, in addition to those pronouns with a special logophoric form (Reuland 2006, p. 3).

Binding Theory

To account for the distribution of n-pronouns vs o-pronouns in Abe, Koopman presented a syntactic analysis of n-pronouns as logical variables, and are Ā-bound by an operator at the Comp. n-pronouns are [+n] and [-n] includes lexical NP, QPs, and O-pronouns, and [+n] elements cannot bind [-n] and vice versa. In logophoric contexts, the complementizer kO is in fact a verb with a [+n] silent subject. The theta role assignment to the subject accounts for the semantic property of logophoric pronouns.

Principle B of Binding Theory states: a pronoun should be free in its domain". This accounts for the disjoint reference between pronouns and possible antecedents within its domain, but not the complex interaction between other possible antecedents. For logophors, it is a subset of all possible interactions that are called logophors.

However, binding theory Principle A is violated in the presence of logophoric reflexives, such as in Icelandic:

In Icelandic, the domain is readjusted to:

Abe

Abe is a Kwa language spoken in the Ivory Coast that uses two classes of third person pronouns: o-pronouns and n-pronouns. O-pronouns behaves like English pronouns, and n-pronouns are used in two different contexts: (1) as a referential pronoun and (2) as a logophoric pronoun. In Abe, the logophoric context must be a combination of: (1) kO-complementation, (2) logophoric verb, and (3) designated argument of the verb.

(1) a. yapii ka api ye Oi,j/n(i),j ye sE
       Yapi tell Api ye he is handsome
    b. yapii hE kO Oj/ni,(j) ye sE
       Yapi said kO he is handsome
       (Koopman 1989: 580 (66))

As (1) shows, not all verbs of saying are logophoric verbs. In addition to its semantic meaning, logophoric verb also depends on the syntactic property of taking on a kO-complement. In addition, only within a subset of kO-complement are logophoric effects observed. There are no logophoric effects observed in (2) regardless of the kO-complementation. The factor is the difference in the source of information, which is relevant to determining what the designated argument is.

(2) a. m hE apii kO Oi,j/ni ye sE
       I said to Api kO she is handsome.
    b. yapii ce kO Oi,j/ni,j ye sE
       Yapi heard kO he is handsome.
       (Koopman 1989: 580 (67))

Icelandic

In Icelandic, reflexives pronouns have two roles: (1) the obligatory clause-bound anaphor, and (2) as a logophoric pronoun. The latter is also referred to as non-clause-bounded reflexives (NCBR) due to its availability to bind with antecedents across multiple clause boundaries:

(1) Formaðurinni varð óskaplega reiður. Tillagan væri avívirðileg. Væri henni beint gegn séri persónulega.
    The-chairmani became furiously angry. The-proposal was(subj.) outrageous. Was(subj) it aimed at selfi personally.
    'It was aimed at him personally, he expressed.'
    (Sells 1987: 453 (26))

The binding of the reflexive anaphor and the antecedent across multiple clause boundaries is restricted to only when the intervening clauses are in subjunctive mood (2b); binding is prohibited across indicative mood as shown in (2a). However, embedding an indicative clause under the verb segja "to say" would allow the subjunctive mood to "trickle" down its structure (3) and allow NCBR:

(2) a. *Joni veit að María elskar sigi
       John knows that Maria loves(ind.) REFL
       'Johni knows that Maria loves himi.'

    b. Joni segir að María elski sigi
       John says that Maria loves(subj.) REFL
       'Johni says that Maria loves himi.'
       (Maling 1984: 212 (2))

(3) Jóni segir að Haraldurj viti að Sigga elski sigi,j.
    Jon says(subj.) that Haraldur knows(subj.) that Sigga loves(subj.) REFL
    Joni says that Haraldurj knows that Sigga loves himi,j.
    (Maling 1984: 223(23b))

The correlation of NCBR and subjunctive mood align with the semantic property of logophoric pronouns. Subjunctive mood is typically the mood of "indirect discourse and other reportive contexts reflecting an individual's point of view". The "trickling down" effect of subjunctive mood when embedded within a verb of saying would allow the reflexive to bind with the subject of the verb of saying.

See also

Anaphora
Bound-variable
Binding Theory
Discoruse
Subjanctive Mood