In their experiment carried out in order to test children’s knowledge of the binding principles, Chien and Wexler (1990) observed that their subjects showed a considerable delay of the acquisition of Principle B compared with Principle A. While the children who participated in their study mastered the locality conditions of anaphors nearly perfectly by the age of six and even earlier, many of them still allowed pronouns to be coindexed with a local antecedent.
In a similar study especially designed to examine children’s knowledge of only Principles B and C, Grimshaw and Rosen (1990) also found that the non-locality constraint for pronouns was to a large extent violated. Like Chien and Wexler, Grimshaw and Rosen state that children yet have innate knowledge of Principle B. A distinction has to be made, they argue, between knowledge of Principle B, on the one hand, and obedience to it on the other. Although children know Principle B, other grammatical and pragmatic influences and circumstances prevent them from obeying it.
Both theses have in common that they are based on the assumption that the delay of Principle B is a universal phenomenon that can be found throughout all languages. However, in a study carried out by McKee (1992), it turned out that the delay of Principle B only affected the group of English children while the Italian children showed nearly perfect mastery of both Principle A and B. McKee accompanies these results by her proposal for an explanation of English children’s delay of Principle B. Since the experiment was carried out with Italian sentences containing clitic pronouns, her account is based on syntactical differences between this type of sentences and its English counterparts containing full pronouns.
This paper will start with a presentation of McKee’s thesis. The presentation will be followed by a discussion in which McKee’s account will be compared with alternative explanations for the delay of Principle B. Since McKee’s syntactically motivated model stand in contrast to other theses based on pragmatics, the final aim will be a judgment about the adequacy and plausibility of both types of approaches. This judgment will be supported by another thesis developed by Cardinaletti and Starke (1995) in which they seek to explain the delay of Principle B by an ambiguity of English pronouns comprising all parts of grammar.
Table of Contents
1. Introduction
2. McKee’s account of English children’s delay of Principle B
3. Other approaches towards an explanation of the delay of Principle B
3.1. Pragmatic constraints
3.2. Knowledge and Obedience
3.3. Ambiguity of pronouns
4. Discussion
4.1. McKee and Chien and Wexler
4.2. McKee and Grimshaw and Rosen
4.3. McKee and Cardinaletti and Starke
4.4. Strong pronouns and emphatic pronouns
4.5. Different GCs for pronouns and anaphors
5. Conclusion
6. Bibliography
1. Introduction
In their experiment carried out in order to test children’s knowledge of the binding principles, Chien and Wexler (1990) observed that their subjects showed a considerable delay of the acquisition of Principle B compared with Principle A. While the children who participated in their study mastered the locality conditions of anaphors nearly perfectly by the age of six and even earlier, many of them still allowed pronouns to be coindexed with a local antecedent. Chien and Wexler argue that children do not violate Principle B but rather lack an additional pragmatic constraint that is needed to interpret pronouns correctly.
In a similar study especially designed to examine children’s knowledge of only Principles B and C, Grimshaw and Rosen (1990) also found that the non-locality constraint for pronouns was to a large extent violated. Like Chien and Wexler, Grimshaw and Rosen state that children yet have innate knowledge of Principle B. A distinction has to be made, they argue, between knowledge of Principle B, on the one hand, and obedience to it on the other. Although children know Principle B, other grammatical and pragmatic influences and circumstances prevent them from obeying it.
Both theses have in common that they are based on the implicit assumption that the delay of Principle B is a universal phenomenon that can be found throughout all languages. This assumption may have been made because Principle B is generally considered to be part of Universal Grammar (UG). However, the universality of the delay of Principle B became controversial after McKee published her study (1992) in which she presented the results of an experiment she had carried out on the knowledge of the binding principles of English- and Italian-speaking children. It turned out as the main result that the delay of Principle B only affected the group of English-speaking children while the Italian children showed nearly perfect mastery of both Principle A and B. McKee accompanies these results by her proposal for an explanation of English-speaking children’s delay of Principle B. Since the experiment was carried out with Italian sentences containing clitic pronouns, her account is based on syntactical differences between this type of sentences and its English counterparts containing full pronouns.
This paper will start with a presentation of McKee’s thesis. The presentation will be followed by a discussion in which McKee’s account will be compared with alternative explanations for the delay of Principle B. Since McKee’s syntactically motivated model stand in contrast to other theses based on pragmatics, the final aim will be a judgment about the adequacy and plausibility of both types of approaches. This judgment will be supported by another thesis developed by Cardinaletti and Starke (1995) in which they seek to explain the delay of Principle B by an ambiguity of English pronouns comprising all parts of grammar.
2. McKee’s account of English children’s delay of Principle B
McKee conducted her experiment by employing the truth value judgment task. The children were presented an event staged with puppets and given a sentence describing the scene. Their task was then to judge whether the sentence describes either correctly or incorrectly what was staged. Since the object of every sentence was realized either by a pronoun or an anaphor, the results of the tests revealed the children’s ability to apply their knowledge of both Principle A and B.
The most strikingly different results between the English- and Italian-speaking children were yielded by the group of one-clause sentences containing pronouns like in (1).
(1a) Lo gnomo lo lava.
The gnome him washes
(1b) Smurfette washed her.
While 85 percent or 100 out of 117 Italian children correctly rejected a sentence like (1a) when given to describe a scene in which the subject performed the action on itself, the sentence (1b) was rejected only by 18 percent or 22 out of 120 English-speaking children when combined with such a scene to result in a violation of Principle B. Since 88 percent or 105 out of 120 English-speaking English rejected one-clause sentences containing an anaphor when used incorrectly and violating Principle A, these results clearly confirm the existence of a developmental delay of Principle B among English-speaking children. Of the Italian children, 94 percent or 113 out of 120 subjects rejected the incorrect usage of one-clause sentences containing anaphors. Despite the lag of 9 percent correct rejections between Italian sentences violating Principle A and B, there exists no delay in the acquisition of one of the two binding principles when this difference is compared with the difference of 70 percent for the English sentences.
These results suggest that some feature or property distinguishing English from Italian causes the delay of Principle B while Italian children are able to demonstrate its knowledge of it at the same age at which they are able to demonstrate knowledge of Principle A. This in turn entails that the search for the difference between Principles A and B causing the delay of the latter can be abandoned because the results of the Italian children show that such a difference does not seem to exist. McKee (1990: 42ff.) rejects earlier attempts to account for the delay of Principle B which were based on the assumption of the existence of such distinct properties. None of these hypotheses, she claims, is able to account for the different performances of English- and Italian-speaking children on the interpretation of pronouns.
In her own proposal, McKee relies on modifications of the definition of governing category (GC) and opposes the standard definition (2).
(2) The governing category for A is the minimal domain containing it, its governor and an accessible subject/SUBJECT.
(Haegeman 1994: p. 241)
Instead, McKee adopts definition (3) of GC suggested by Huang (1983), which, according to him, should be applied while the binding principles for anaphors and pronouns should remain unchanged.
(3) α is a governing category for β if and only if α is the minimal category containing β, a governor of β, and a SUBJECT that, if β is an anaphor, is accessible to β.
(Huang 1983: p. 557)
McKee claims that (3) correctly defines the GC used by both the Italian- and English-speaking children who participated in her experiment. This definition means that GC as defined by (2) only applies to anaphors while pronouns need to be unbound in a GC constituted by the minimal category containing the pronoun itself as well as its governor but no subject in any sense.
While both groups of children, according to McKee, obey Principle B and, furthermore, use an identical definition of GC for pronouns, the crucial point that leads to the different results in the interpretation of pronouns sentences is the difference between English sentences containing full pronouns and Italian sentences containing clitic pronouns. The syntactical position of these proforms is still subject to discussion but McKee follows the assumption according to which a clitic pronoun is part of INFL. The structural representation of (1a) is (4a) in comparison to (1b) with a full pronoun represented in (4b).
For (4a), the only possible GC for the clitic pronoun is IP since it not only is the maximal projection dominating INFL and the clitic but at the same time constitutes the sentence. For (4b), the situation is more ambiguous since the pronoun is, as the complement NP of VP, dominated by two maximal projections, VP and IP. With definition (2) applied to (4b), the pronoun’s GC would be IP since the VP only contains the pronouns itself and V as its governor but not a subject/SUBJECT for which either the subject NP Smurfette or the INFL with AGR apply both of which are immediately dominated by the IP node. With definition (3), on the other hand, the pronoun would only have to be free in the VP because it contains the pronoun and the governor and hence, according to this definition, all elements required to form the GC.
The GC in sentence (4b) as generated by the application of definition (3) is claimed responsible for the weak performances of the English-speaking children. While the Italian children correctly reject a context in which the pronoun is coindexed with the subject NP, such a context is not rejected by the English-speaking children because the subject NP is not contained within their assumed GC. The pronoun is allowed to be bound by the subject NP without violating Principle B. To sum it up: Both groups of children obey Principle B and form the GC for pronouns according to definition (3). The different syntactical features of full pronouns, on the one hand, and clitic pronouns on the other are the origin of the significantly different performances of the two groups in the interpretation of sentences containing a pronoun.
3. Other approaches towards an explanation of the delay of Principle B
3.1. Pragmatic constraints
As a result of their experiment carried out solely on English-speaking children, Chien and Wexler (1990) also observed that their subjects performed worse on the interpretation of pronouns than on sentences containing anaphors. This delay can be explained, they argue, when not only the purely syntactical Principles A and B are taken into consideration but also semantics of pronouns and anaphors. Their interpretation, then, is governed by a pragmatic principle.
Chien and Wexler state that noncoindexed NPs do not necessarily have to be disjoint but can be coreferential in certain context. They demonstrate their thesis on (5).
(5a) That must be John.
At least he looks like him.
(5b) That must be John.
* At least hei looks like himi.
(5c) That must be John.
At least hei looks like himj.
(Chien and Wexler 1990: p. 256)
According to Principle B, the second sentence in (5b) is ungrammatical because him must not be coindexed with its local c-commanding antecedent he. Nevertheless, despite the noncoindexing in (5c), the two NPs are corefential in this context. While the purely syntactical Principle A also implies semantic coreference for anaphors, this connection does not hold for pronouns because syntactic noncoindexation does not imply semantic disjointness. Principle A determines a NP’s reference but Principle B does not entail such a determination. Noncoindexed NPs refer freely and thus may or may not corefer.
However, when the second sentence of (5) occurs without any context as in (6), the two NPs are generally understood as not being coreferential.
(6a) He looks like him.
(6b) * Hei looks like himi.
(6c) Hei looks like himj.
(Chien and Wexler 1990: p. 257)
[...]
-
-
-
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X.