The Pragmatic Effect on Binding in the English Language. Influence of Person and Pronoun

Scientific Study, 2019

39 Pages, Grade: 1,3











“[R]esearchers have seen binding phenomena as a rich source of data about the universals of human language.”

Featherston & Sternefeld (2003: 1)

Featherson and Sternefeld (2003) point out the significance of binding: Binding is not only a subarea of linguistics, as it provides more information than expected. The binding phenomena reveals several other aspects of the human language, such as its complexity and its dynamic interaction of syntax and context. Since the pioneering Pisa Lectures on Government and Binding by Noam Chomsky (1981, 1986), it has commonly been assumed that a single theoretical apparatus, such as the Binding Theory universally determines the interpretation and distribution of pronouns and anaphors in English. The Binding Theory proposed by Chomsky expresses a generalization about syntactic relationships by using three principles. This has been criticized, refined and modified over the last decades. In particular, the revision proposed by Moskovsky (2004) attempts to provide remarks and improvements for the existing version of Chomsky’s classic Binding Theory. This is the reason why we were interested in doing more research on binding. We wanted to determine whether Moskovsky’s criticism can be supported or not. Therefore, an experiment on binding was conducted.

In the following paper, I will first provide some theoretical background information on binding in English. The emphases here will be especially on Chomsky’s classic Binding Theory (1988) and Moskovsky’s criticism. I will then explain the motivation and design of our experiment by presenting variables, conditions, and items. In the following, the results of the experiment will be presented and interpreted. A final comparison to the Turkish language will further argue whether binding can be considered to be a universal phenomenon or not.


In this section, I will illustrate Moskovsky’s criticism on Chomsky’s classic Binding Theory. Therefore, the basic principle of Binding Theory will be clarified first.

The classic Binding Theory describes the conditions on the structural relations between nouns and therefore, it deals with three specific types of noun phrases (NPs), namely R-expressions, pronouns and anaphors. Chomsky’s classic Binding Theory consists of the following three principles as in (1):


(A) An anaphor is bound in its local domain.
(B) A pronominal is free in its local domain.
(C) An R-expression is free. (Chomsky 1988:188)

A local domain, which can be also called “binding domain”, is the smallest IP (Inflectional Phrase) containing the NP (Noun Phrase). For the case of Principle A of the Binding Theory, the binding domain is the smallest IP containing the anaphor and its antecedent. I will explain this more precisely by using three example sentences:

(2) a. [IP Lillyj danced with herselfj ].

b. [IP Cameroni said [CP that [IP Lillyj danced with herselfj ] ]].
c. *[IP Cameroni said [CP that [IP Lillyj danced with himselfi ] ]].

The smallest IPs are marked by bold squared brackets. In (2a), Herself is the anaphor which has the antecedent Lilly within its smallest IP. In sentences such as (2b), which contain two IPs, the embedded clause [IP Lillyj danced with herselfj ] is the binding domain of the anaphor. The anaphor herself has its binder Lilly within this embedded IP. Hence, sentences (2a) and (2b) are grammatical. Whereas (2a) and (2b) are grammatical, sentence (2c) is ungrammatical because the anaphor is not bound within its binding domain. Himself refers to the antecedent Cameron which is outside the smallest IP. This violates Principle A of the Binding Theory. Therefore, sentence (2c) is not grammatical.

Principle B of the Binding Theory says the opposite to Principle A. The crucial difference to an anaphor is that a pronoun is not bound within its local domain. I will show you three example sentences below. As in the previous examples, the smallest IPs (binding domains) are marked by bold squared brackets. Sentence (3a) is grammatical, as the pronoun him does not refer to the subject Joey which means that the pronoun is free in its binding domain. This is shown by the different indexes j and k. Hence, the binder of the pronoun him must be outside the sentence (3a). The same pattern works for sentence (3b). (3b) has two IPs and the smallest IP containing the pronoun is the embedded clause [IP Joeyj hit himi ]. The antecedent John of the pronoun him is outside the binding domain which means that the pronoun is free. In contrast to (3a) and (3b), the last sentence (3c) is not grammatical, as it violates Principle B. The pronoun him in the inner IP has its binder Joey within the binding domain. As already mentioned, this condition solely applies to Principle A of the Binding Theory.

(3) a. [IP Joeyj hit himk ].

b. [IP Johni said [CP that [IP Joeyj hit himi ] ]].

c. *[IP Johni said [CP that [IP Joeyj hit himj ] ]].

Based on this knowledge, we can have a closer look at Principles A, B and C. Principle A of the Binding Theory claims that a noun phrase (antecedent) X must bind an anaphor Y in its local domain. As already explained, the local domain here is the smallest clause containing the anaphor Y. X binding Y is fulfilled, if and only if X and Y are co-indexed and X c-commands Y (Featherston 2018). Consider the following example:

Abbildung in dieser Leseprobe nicht enthalten

(4) Johni hit himselfi.

Sentence (4) contains the NP John (antecedent) and the anaphor himself. The tree below the sentence shows clearly that both criteria of principle A are fulfilled. The antecedent John c-commands the anaphor himself. A c-command, i.e. constituent-command, depicts the relationship of dominance between the nodes of a grammatical tree, as in (2). A node X c-commands the node Y only if a sister of X dominates Y. Sisters in a linguistic structural relation (trees) are sibling nodes which have the same mother. In our case, the sister of the DP John is , as John and have the same mother IP. Accordingly, John c-commands himself, since himself is dominated by the sister node of John. In addition to the c-command, the second condition of Principle A is fulfilled. The indices illustrate that John and himself belong together, hence, are co-indexed.

In contrast to that, Principle B of the Binding Theory says that a pronoun must be free. In other words, pronouns are not bound within the local domain (Chomsky 1981). Consider the following example sentences:

(5) a. Johni hates himj. (him ≠ John)

Abbildung in dieser Leseprobe nicht enthalten

b. *Johni hates himi.

Abbildung in dieser Leseprobe nicht enthalten

Though the antecedent John does c-command the pronoun him in the tree above, the sentence (5a) shows that him is still free in its local domain as there is no co-indexation between the pronoun him and the antecedent John. Both NPs not being co-indexed clearly implies that him does not refer to John in this sentence. The pronoun needs to refer to another 3rd person. However, according to principle B, the sentence (5b), which is structurally identical with (5a), would be unacceptable. A co-indexation in sentences like this would indicate that a pronoun is bound within the local domain. More precisely, the co-indexation would imply that the pronoun him is the referent to the antecedent John. The co-indexation, however, is restricted to be solely a criterion of principle A. As both sentences (5a) and (5b) only differ in matters of co-indexation, it becomes clear that the unacceptability of (5b) cannot be accounted for in terms of syntactic structure, but rather from a non-structural (pragmatic) factor here (Moskovsky 2004). If the co-indexation was left out, both sentences would be structurally identical. The next paragraphs will discuss this issue more precisely as being part of Moskovsky’s investigations on third person binding.

Before proceeding to Moskovsky’s investigations on third person binding, I will present the last condition of the classical Binding Theory: Principle C. This principle says that an R-expression is globally free. In other words, R-expressions are neither bound, nor co-indexed, nor c-commanded by anything it refers to (Chomsky 1981). R-expressions are referential expressions which express content and get their meaning by referring to entities in the real world. Unlike anaphora, reciprocals, or pronouns, R-expressions are descriptively richer. Some examples for R-expressions could be common nouns such as “an iPad”, or proper names such as “Lucy” or “Mitchell”. Consider the following two example sentences:

(6) a. Hej admires Mitchelli.

b. *Hei admires Mitchelli.

Abbildung in dieser Leseprobe nicht enthalten

Mitchell is he R-expression in both sentences. The first sentence (6a) is grammatical as Mitchell is neither bound, co-indexed or c-commanded to anything it refers to. In contrast to that, the R-expression Mitchell in sentence (6b) is bound to the antecedent He within the sentence. As Mitchell is not free in this example sentence, it violates Principle C and becomes ungrammatical.

There are several criticisms and suggestions for improvement on Chomsky’s classical Binding Theory. One of these improvements is suggested by Moskovsky’s (2004). In general, it examines problems of Chomsky’s classic Binding Theory in sentences which are differing in morphological number, though having the same structure (Moskovsky 2004). Chomsky assumes that binding depends on syntactical factors, whereas Moskovsky’s “Third person effects on binding” is based on the assumption that binding is subject to pragmatic factors (Moskovsky 2004). According to this hypothesis, sentences with the 1st and 2nd person do allow the binding of pronouns and only sentences with the 3rd person which do not allow pronominal binding. Although this assumption violates the existing version of Binding Theory, Moskovsky provides the following two example sentences as evidence:

(7) a. I am not thinking of me.

b. *She i is not thinking of her i . (Moskovsky 2004:1035)

Though both sentences (7a) and (7b) are structurally identical, Moskovsky argues solely for the acceptability of the sentence (7a). According to “Third person effects on binding”, the use of the 1st person in the sentence (7a) allows the binding of the pronoun to be acceptable, whereas the latter does not allow any pronominal binding because it contains the 3rd person. What can be said so far is that the principle B (pronominal must be free) of the classic Binding Theory cannot explain the acceptability of (7a). What we need here as an explanation is the Neo-Gricean Conversational Principle which Moskovsky calls Avoid Ambiguity Principle here (2004) . This principle states that if there are two syntactically identical types A and B of a language, and structure A is ambiguous between the meanings X and Y, and structure B has only meaning X, speakers of this language should use structure A to express meaning Y and B to express X (Dowty 1980 qtd. in Moskovsky 2004). In brief, the sentence (7a) is not ambiguous, as the pronoun me can only refer to the subject of the sentence, namely I. On the other hand, the pronoun her in the sentence (7b) can refer to any 3rd person. This may be the subject she or any other referent. This is why Moskovsky revises the classic Binding Theory by proposing the usage of the Avoid Ambiguity Principle in combination with the Binding Theory for 3rd person sentences.

So let us recap what we know so far. Moskovsky criticizes Chomsky’s classic Binding Theory and proposes the operation of the Avoid Ambiguity Principle for 3rd person sentences as in (8):

(8) a. *Hei saw himi.

(syntactic structure A = ambiguous between X and Y)

b. Hei saw himselfi.

(syntactic structure B = unambiguous = only meaning X)

Syntactically, (8a) and (8b) are sentences of the equally simple type A and B. Following the Avoid Ambiguity Principle, the first sentence is structure A. It is ambiguous, as him can refer to any 3rd person. This could be the antecedent he or any other referent. On the other hand, there is the sentence (8b) which is an example for structure B. It is unambiguous since the reflexive pronoun himself can only refer to the subject he. When the intended referent is someone outside of the sentence, speakers should use sentences like (8a) to avoid any confusion. But if the referent and the subject of the sentence are the same person, speakers should use sentences like (8b).

To sum up, Moskovsky (2004) attempts to provide a solution for a problem of the classic Binding Theory. There are several crucial points that Moskovsky claims: unambiguous structures are always preferred over ambiguous ones. Since reflexive pronouns are always unambiguous, they are mostly preferred over pronouns where there is ambiguity. The 1st and 2nd person pronouns are also unambiguous. They can normally refer to one specific person, namely the speaker or hearer of the sentence, whereas the 3rd person can have more than one possible referent. Consequently, the question is whether 1st and 2nd person pronouns are as good as reflexives.

In order to investigate Moskovksy’s assumption, an experimental case study was conducted. This will be presented in the following section.




The goal of this experiment is to find out whether the distinction between the 1st, the 2nd and the 3rd person is as clear as Moskovsky (2004) argues and whether there is a difference between the use of personal pronouns and reflexive pronouns. Therefore, we used transitive verbs and distinguished between 1st, 2nd and 3rd person both singular and plural, as well as we distinguished between personal pronouns, reflexive pronouns and names. The distinction between personal pronouns and reflexive pronouns is necessary as the point of this experiment is whether 1st and 2nd personal pronouns are equal to reflexives.

Transitive verbs need a complement or an object directly following it to make the sentence meaningful. In our experiment, the subject and the object of every single transitive verb refer to the same person. The following sentences will exemplify transitive verbs and the distinction between the 1st person and the 3rd person:

(9) a. Ii am looking at mei.

b. *Hei is looking at himi.

The verb look at needs two syntactic arguments, namely someone who looks and something or someone that is looked at. Therefore, the NP me is an obligatory argument, which contributes to the meaning of the verb phrase. As the focus of this experiment is on differences between pronoun and grammatical person, the arguments in (9a) and (9b) refer to the same person used in each sentence. In (9a) it is clear that the subject pronoun I refers to the object personal pronoun me. In the case of the 1st person, there is no ambiguity since it can only refer to one specific person. However, in sentences like (9b), this is not the case. 3rd person pronouns can have more than one referent. Thus, it is not clear whether the object personal pronoun him refers to the subject pronoun he or to another third male person. Here, the index i indicates that him is meant to refer to the subject he which makes (9b) ungrammatical. If the index i was left out, the sentence in (9b) would be grammatical as the object pronoun would not refer to the subject pronoun.

(10) a. *Johni saw himi.

b. Johni saw himselfi.

As mentioned before, 3rd person pronouns can have more than one possible referent. Both (9b) and (10a) are ungrammatical for the same reason. In both sentences, it is not clear whether the object personal pronoun him refers to the subject pronoun He or John or to another third male person. Here, the index i indicates that him is meant to refer to the subject he which makes (9b) and (10a) ungrammatical. This violates Principle B of the Binding Theory. If the index i was left out in both sentences, the pronouns would be free in their binding domains and Principle B would apply properly. As a result, both (9b) and (10a) would be grammatical. This is the reason why 3rd person sentences need reflexives as in (10b).

This leads to the conclusion that we need two variables to compare both the influence of grammatical person and pronoun type on binding. The first variable, which is depicted as V1 in the table below, is grammatical person with the values containing person and grammatical number. This means it starts with the 1st person singular and goes on with the 2nd person singular, 3rd person singular, 1st person plural, 2nd person plural and 3rd person plural. The second variable V2 contrasts personal pronouns and reflexive pronouns. As a result, the combination of these variables gives us only twelve conditions. In order to create 16 conditions, four more conditions are needed. We add four “dummy” conditions, namely the value “name” which gives us the last four conditions marked with a cross in the table.

Table 1: CONDITIONS – Variable 1: grammatical person, Variable 2: pronoun

V1: grammatical person

Abbildung in dieser Leseprobe nicht enthalten


Now that all variables and conditions were set up, we needed to create our items. Therefore, we formed eight natural sounding sentences by using eight reflexive verbs which are the base forms for our conditions. The items were created in each of the conditions, therefore, the experimental material included 128 items in complete. The full set of materials can be found in the appendix at the end of this paper. One set of 16 conditions will be shown below to give an idea of the remaining 112 items’ structure:

Table 2: ITEMS – one set of sample items with 16 conditions

Abbildung in dieser Leseprobe nicht enthalten


Excerpt out of 39 pages


The Pragmatic Effect on Binding in the English Language. Influence of Person and Pronoun
University of Tubingen
Catalog Number
ISBN (eBook)
ISBN (Book)
binding, effect, english, influence, person, pragmatic, pronoun
Quote paper
Anonymous, 2019, The Pragmatic Effect on Binding in the English Language. Influence of Person and Pronoun, Munich, GRIN Verlag,


  • No comments yet.
Read the ebook
Title: The Pragmatic Effect on Binding in the English Language. Influence of Person and Pronoun

Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free