Imposters: A Study of Pronominal Agreement

15. Fake Indexicals

15 Fake Indexicals 15.1 Partee’s Observation An observation by Partee (1989) initiated a discussion of so-called fake indexicals : forms that have the shapes of non–3rd person pronominals but are interpreted as bound variables (for careful discussions, see Kratzer 1998, 2009; Rullmann 2004; Heim 2008). The following data illustrate a 1st person singular fake-indexical pronoun: (1) a. I think that I know how to ski well. b. Only I think that I know how to ski well. Unlike (1a), (1b) is ambiguous. The standard reading implies that no person other than the speaker thinks that the speaker knows how to ski well. The other, bound variable reading is interpreted as ‘no x other than the speaker thinks that x knows how to ski well’. Examples (2a,b) admit the same ambiguity. (2) a. I am the only one who thinks I know how to ski well. b. It is only me who thinks I know how to ski well. The notable property in both (1b) and (2) is the existence of a bound variable reading for I (we will mostly ignore the fact that the analogous situation holds for you). A form with the standard shape of a 1st person pronominal that is not interpreted as referring to the speaker is called a fake indexical. The normal 1st person forms are real indexicals. This can be made more precise in present terms as follows: (3) A non–3rd person pronominal P is a fake indexical if and only if P’s ultimate antecedent is neither AUTHOR nor ADDRESSEE. Examples such as (2a) fall under our concept of secondary sources. Relevant in particular is the specification rendering the subject of a predicate nominal a source for anything the predicate nominal is a source for. An analysis of (2a) in 192 Chapter 15 these terms is sketched in (4). Despite significant differences, the following account is in the spirit of Heim 2008 and Kratzer 2009. (4) In (2a): a. The REL DP who is the immediate antecedent of I. b. The DP [the only one who . . . ] shares a lexical basis with the REL DP who and hence is a source for anything that who is a source for. c. The matrix I is the subject of the predicate nominal DP [the only one . . . ]. d. Hence, via our source schema in (3) of chapter 13, the matrix I is a secondary source for the embedded I. e. AUTHOR is the immediate and ultimate antecedent of the matrix I. Under this analysis, the 1st person value of the embedded I is attributed to agreement with a secondary source, here AUTHOR, as permitted by the Pronominal Agreement Condition (4) of chapter 14. 15.2 Only I The ambiguity of examples like (1b) becomes comprehensible once one highlights the fact that a DP like only I represents two distinct 1st person singular DPs (on the ϕ-feature values of only I, see Heim 2008, 48). Since the 1st person singular status of the DP I is not controversial, the overall claim is supported by the observation that only I is also 1st person singular, as shown by verb agreement. (5) Only I am/*is aware of his origins. Moreover, in a parallel fashion, only you and only we each represent two distinct DPs, each with the same person and number values. In such cases, it appears that the larger DP agrees with its component DP. This is an uncommon situation, since, for example, the complex DPs in (6) reveal no agreement between dominating and dominated DP. (6) a. No one but I/me *am/is aware of his origins. b. No attack on me *am/is going to change my mind. c. No one like you *are/is available. The agreement of only DPs with their component DP may well motivate a further extension of the notion primary source in (22) of chapter 14. To state such an extension precisely would require some kind of characterization of only DPs that distinguishes them in general terms from DPs like those in (6a–c). We will not attempt such a characterization here. We note, though, that Fake Indexicals 193 it could not be limited to only forms, since the relevant agreement property is also found with expressions of the form even DP and just DP (DPs modified by focus particles). Returning to the ambiguity of (1b), one can simply say that the standard...