Jump to ContentJump to Main Navigation
ImpostersA Study of Pronominal Agreement$

Chris Collins and Paul M. Postal

Print publication date: 2012

Print ISBN-13: 9780262016889

Published to MIT Press Scholarship Online: August 2013

DOI: 10.7551/mitpress/9780262016889.001.0001

Show Summary Details
Page of

PRINTED FROM MIT PRESS SCHOLARSHIP ONLINE (www.mitpress.universitypressscholarship.com). (c) Copyright The MIT Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in MITSO for personal use (for details see www.mitpress.universitypressscholarship.com/page/privacy-policy). Subscriber: null; date: 21 January 2019

Crosslinguistic Variation in Pronominal Agreement

Crosslinguistic Variation in Pronominal Agreement

(p.225) 19 Crosslinguistic Variation in Pronominal Agreement

Chris Collins

Paul M. Postal

The MIT Press

Abstract and Keywords

This chapter reviews some crosslinguistic evidence that shows how pronominals agree with imposters. It suggests that languages divide into two groups depending on how pronominal agreement with imposters works. Specifically, it proposes a parameter for pronominal agreement in which an imposter A immediately antecedes a pronominal P, focusing on examples from Mandarin and Indonesian. It then compares Mandarin and Indonesian imposters with ass-camouflage construction determiner phrases.

Keywords:   pronominals, imposters, languages, pronominal agreement, Mandarin, Indonesian, ass-camouflage construction, determiner phrases

19.1 Proposed Parameter

This chapter is a very preliminary report on some crosslinguistic data we have accumulated about how pronominals agree with imposters. Given this information, we speculate that languages divide into two groups depending on how pronominal agreement with imposters works. Specifically, we propose the parameter of variation in (1). (We return to an alternative in section 19.4.)

  • (1) The Pronominal Agreement Parameter

    If an imposter A immediately antecedes a pronominal P, then:

    • a. P agrees only with A.

    • b. P agrees only with A’s ultimate antecedent.

      (arguably Mandarin Chinese, Indonesian)

    • c. P agrees with A or A’s ultimate antecedent.

      (English, French, Bellinzonese)

The following sections support the suggested distinctions.

19.2 Mandarin and Indonesian

Wang (2009, 3) provides data like (2) and (3) showing that Mandarin imposters only give rise to non-3rd person pronominal agreement.

  • (2) Laoshi kuai yao  shiqu (wo/*ta) de naixing le.

    teacher almost going.to lose I/3 SG POSS patience INCHOATIVE

    ‘Teacher (= I) is going to lose (my/*his/*her) patience.’

  • (3) Benren  yijing jueding (wo/*ta) bu chu guo  le.

    this.person already decide I/3 SG  NEG exit country INCHOATIVE

    ‘This person (= I) decided that I/*he would not go abroad.’

(p.226) Wang documents that singular and plural imposters anteceding local reflexives, long-distance reflexives, and the controlled elements of control clauses all obey the same generalization: the agreeing form must match the ultimate antecedent, never the imposter shell.

Kaufman (2009, 4–5) reports a similar set of facts for Indonesian.

  • (4) Pak gurui sudah mem-baca koran itu sebelum andai pergi.

    Mr. teacher already Av-read paper that before 2SG go

    ‘Teacher (= you) already read the newspaper before you left.’

  • (5) Pak gurui sudah memperkenalkan diri andai.

    Mr. teacher already introduced  self 2SG

    ‘Teacher(= you) already introduced yourself.’

In these data, when an imposter is anteceded by ADDRESSEE, it must antecede a 2nd person pronoun. When Pak guru ‘Mr. Teacher’ antecedes a 3rd person pronoun, the imposter interpretation is not available.

Kaufman (2009) reports on a significant twist distinguishing the Indonesian and Mandarin data. Indonesian 1st and 2nd person pronouns representing the agent are able to appear in a preverbal clitic position in the patient voice; see (6). Nonimposter, nonpronoinal agent DPs cannot occur in this position; see (7).

  • (6) Filem itu sudah kau/kita/kalian=tonton.

    film that already 2SG/1PL/2PL=watch

    ‘That film was already watched by us/you.’

  • (7) *Koran itu sudah [pak gurui]=baca sebelum diai pergi.

    paper that already Mr. teacher=read before 3SG go

    ‘The newspaper already was read by Teacher before he left.’

However, when a DP is interpreted as a 2nd person imposter, it can appear in preverbal position.

  • (8) Koran itu sudah [pak gurui]=baca sebelum andai pergi.

    paper that already Mr. teacher=read before 2SG go

    ‘The newspaper was read by Teacher before you left.’

The imposter interpretation in (8) is forced by the use of the 2nd person pronoun in the subordinate ‘before’ clause. When the DP ‘Mr. Teacher’ appears postverbally, the sentence with a 2nd person pronoun in the adverbial clause is ungrammatical.

  • (9) Koran itu sudah di-baca (oleh) pak gurui sebelum dia1/*anda1 pergi. paper that already PV-read by Mr. teacher before 3SG/2SG go

    ‘The newspaper was read by Teacher before he left.’

(p.227) The data in (6)–(9) provide striking support for our basic syntactic analysis of imposters. That is, that analysis permits the following otherwise unavailable generalization:

  • (10) A DP appears in clitic preverbal position only if it has a non–3rd person ultimate antecedent.

19.3 Comparison with the ACC

Mandarin and Indonesian imposters behave similarly to ACC DPs. Recall, from chapter 6, that SHCC DPs show ambivalent pronominal agreement.

  • (11) Your Majesty should take better care of herself/yourself.

However, as discussed in chapter 17, such variation is impossible with ACC DPs (Collins, Moody, and Postal 2008, 39).

  • (12)

    • a. Your ass making a fool of yourself/*itself/*hisself

    • b. Your ass was late, wasn’t you/*he/*she/*it?

In this respect, then, Mandarin and Indonesian imposters behave like English ACC DPs, and English imposters behave like English SHCC DPs, as summarized in (13).

  • (13)

    Pronominal agreement

    With ultimate antecedent

    With immediate antecedent

    English ACC



    Mandarin imposters



    Indonesian imposters



    English SHCC



    English imposters






These highly preliminary data raise three crosslinguistic research questions. First, a logically possible third type of language is one where the pronoun only agrees with the shell. Do such languages exist? Second, we have only described pronominal agreement with imposters, which make the ultimate antecedent available as a secondary source for the pronoun. Do the facts of Mandarin and Indonesian hold for pronominal agreement involving other secondary sources described in chapter 13? Third, does the behavior of Mandarin and Indonesian imposters correlate with other aspects of those languages (e.g., the analysis of (p.228) pronouns or something about the internal syntax of their imposters)? Such questions, which could not be formulated previously, can now form the basis of a crosslinguistic research agenda for imposters. See Das 2009 for some preliminary discussion of Bengali in terms of these questions.

19.4 An Alternative

An alternative way of characterizing the data from Mandarin Chinese and Indonesian would deny that these languages even have imposters. On this alternative, the explanation of (2) would claim that ‘teacher’ is embedded in an appositive (precursor) structure with a covert pronoun. So laoshi ‘teacher’ in (2) would actually be [wo, laoshi] with a covert 1st person singular pronoun (this analysis was proposed in Wang 2009). Such an account would handle the above data equally well.

The reason why the 3rd person singular pronoun is unacceptable in (2) is that its antecedent is the 1st person DP [wo, laoshi]. On this alternative analysis, the correct parameter distinguishing English from Mandarin and Indonesian is (14).

  • (14) The Imposter Parameter

    • a. L has imposters.

      (English, French, Bellinzonese)

    • b. L does not have imposters.

      (arguably Mandarin, Indonesian)

Under this analysis, one more parameter of variation is needed: whether or not the pronoun in an appositive (precursor) can be null. In Indonesian and Mandarin, it would have to be the case that the pronoun is null.