1 Introduction
English demonstratives and, in particular, the choice between the proximal demonstrative this and the distal demonstrative that, have received a huge amount of attention in the literature.Footnote 2 As is well known, English demonstratives can either be determiners in an NP, as in (1a), or function alone in an NP, as in (1b), traditionally called the pronominal use of the demonstrative:Footnote 3
(1)
(a) This book is the best I've ever read.
(b) This is the best book I've ever read.
Following a long tradition, we will call the underlined NPs in (1) demonstrative NPs.
Classically, three uses of demonstrative NPs have been distinguished. First, the deictic use, where the referent is physically copresent with the speaker and addressee (see, e.g. Clark & Marshall Reference Clark, Marshall, Webber, Joshi and Sag1981), often accompanied by some form of pointing. In the following example, speaker and addressee are in a museum, standing in front of the painting being referred to, which has not been mentioned previously:
(2) So this is one of the last paintings Van Gogh ever painted. (Dr Who, S5, Ep10)
In this occurrence, there is no manual pointing (though of course the speaker's gaze is directed towards the referent). Note that the predicative complement (with the head noun paintings) plays a crucial role in making the intended referent unambiguous. Other predicates could lead to different referent choices in the same context without any necessary change in the direction of the gaze (e.g. So this is the original frame / the end of our tour / the last room we will be visiting; see Webber Reference Webber1988: 5).
Second, the anaphoric use, where the referent is recovered via an antecedent present in the co-text, i.e. a linguistically co-present discourse segment (see Clark & Marshall Reference Clark, Marshall, Webber, Joshi and Sag1981):Footnote 4
(3) When you and I are in a room, that room becomes charged, absolutely electric. (All My Children, 2010-03-30, SOAP)
Many references make a distinction between cases like (3), where the antecedent is an NP, and cases where the antecedent (or more precisely, the antecedent-trigger; see Cornish Reference Cornish1999: 41ff.) is a more complex discourse segment. These have been called ‘discourse deixis’ (Diessel Reference Diessel1999: 100ff.; other denominations include ‘impure textual deixis’, Lyons Reference Lyons1977: 668, and ‘anadeixis’, Cornish Reference Cornish, Nilsen, Appiah Amfo and Borthen2007).
Third, the recognitional use (see e.g. Diessel Reference Diessel1999: 105ff.) where the referent is neither physically nor linguistically co-present, but is rather part of shared knowledge, accessible through previous interactions between the interlocutors and their shared community membership (see Clark & Marshall Reference Clark, Marshall, Webber, Joshi and Sag1981):
(4) You have a ton of friends, you got married, moved into a new apartment, you wore a baseball hat that one time. (The Big Bang Theory, S.12, E.23)
In what follows we will refer to these three classical uses of demonstrative NPs as involving three ‘modes of reference’, deictic, anaphoric and recognitional respectively.
In this article, we will focus on the choice between proximal this and distal that in deictic uses. Huddleston & Pullum et al. (Reference Huddleston and Pullum2002: 1505) provide a typical account:
The primary use of the demonstratives is in NPs referring to objects present in the situation of utterance, with this applying to objects relatively close to the speaker (proximal), and that to objects relatively distant from the speaker (distal): … What counts as proximal and what as distal is not determined by purely objective features of spatial location: there may be a subjective element involved. For example, I might be holding something in my hand and still have a choice between saying What is this? and What is that? In this context this would be the default choice, but that could be used to indicate some negative attitude such as disapproval.
This ‘subjective’ element has been given much attention in the literature, with very illuminating results; see in particular Lakoff (Reference Lakoff1974), Cornish (Reference Cornish2001), Strauss (Reference Strauss2002) and Peeters, Krahmer & Maes (Reference Peeters, Krahmer and Maes2021) for a recent overview. As the latter point out: ‘the influence of physical factors decreases as a function of an increase of importance of the addressee in the speech situation at hand … and … psychological factors are by default most important in shaping a speaker's choice of demonstrative form in natural, communicative situations’ (Peeters, Krahmer & Maes Reference Peeters, Krahmer and Maes2021: 417). In what follows, I will use the expression ‘psychological/subjective’ to refer to these non-physical factors that influence the choice between proximal and distal.
Consider the following example:
(5) [Krystal:] Would you open the doors? Unlock them right now. […] Would you look at me? Look at me, Adam. You've got your family locked inside that room. Now, I want you to open those doors. (All My Children, 1/18/08, SOAP)
In this example, Adam is closer to the doors and the room on the other side than Krystal, but this physical factor is not what is central to the choice of the distal that and those. The crucial point is that the addressee has locked the doors and is directly concerned with what is going on in the room on the other side. Notice that if the speaker, Krystal, had been just as close to the doors and the room as Adam, but he was the one with his hand on the doorknob, the proximal demonstrative would still have been dispreferred. In the same situation, with speaker and addressee equidistant from the doors and the room, if Adam had been speaking, he would have been more likely to refer to the doors and room using proximal demonstratives:
(6) [Adam:] I've got my family locked inside this room. Now I'm going to open these doors.
Note that even in this case, it is not true that the distal alternative would be impossible. As discussed at length in the very interesting literature on the subject previously cited, and summarized by Peeters, Krahmer & Maes (Reference Peeters, Krahmer and Maes2021), the psychological/subjective factors at play in the choice are multifactorial and linked to a variety of properties of the context and the speaker. Examples (5) and (6) are then of a type similar to that discussed in the above quote from Huddleston & Pullum, where, in the same physical situation, a choice is possible between demonstrative NPs with this and that to refer to the same entity.
It has also been noted in the literature that there is an asymmetry between proximal and distal referents. Contrary to what is the case with referents that are close to the speaker, where both this and that are possible, as suggested by Huddleston & Pullum, when a referent is far from the speaker, the use of the proximal demonstrative is less natural. If the speaker is in a classroom where the windows are on the other side of the room, it would be less felicitous to say Can you open this window? while pointing to one of the windows. Only that window would be felicitous.Footnote 5 The broader range of referential possibilities of that is the central reason for considering it to be the unmarked member of the pair.Footnote 6
In his typology of uses of demonstratives, Levinson (Reference Levinson, Horn and Ward2004: 108) (following Fillmore Reference Fillmore1997: 63) distinguishes, among deictic uses, between gestural deixis and symbolic deixis. Gestural deictic uses require some form of pointing gesture, typically with the hand or the gaze, for reference to succeed. Fillmore provides this finger as an example of this type. Symbolic deixis, on the other hand, involves exophoric reference without requiring pointing. Fillmore gives this campus as an illustration. This article will focus on these cases of symbolic deixis, where pointing is not necessary.
More specifically, it appears not to have been noticed in the literature that in these symbolic deictic uses of demonstrative NPs only this is possible and switching to that would lead to a change in reference. These are the cases which will be the focus of the present article. Specifically I will show that only the proximal demonstrative this is possible when referring deictically to a referent in which the speaker is contained, as opposed to those cases where the speaker is not contained in the referent and the choice between proximal and distal demonstratives involves the classical notion of closeness (physical or psychological/subjective) to the speaker. More precisely, I will argue that there are four relevant types of containment. The first and most obvious case is spatial containment, where the speaker is physically located inside the referent (e.g. this room). Then, there are three derived cases, which can be considered to be metaphorical extensions of the first, namely: situational containment, where the speaker is a participant in an event or state (e.g. this conversation); set containment, where the speaker is a member of the set denoted by the referent (e.g. this family); and temporal containment, where the speaker (or, more precisely, the time of utterance) is temporally contained in the referent (e.g. this week). We will examine these four cases in turn.
2 Spatial containment
When the intended referent is a first-order entity, the speaker is either spatially contained in the entity or not contained in the entity. Let us return to examples (5) and (6). The speaker is outside the room being referred to and is thus not spatially contained in it. As we noted, both the distal and the proximal demonstrative are possible, the choice between the two being explained in terms of psychological/subjective factors.
Consider now the variant of those examples where the speaker and addressee are inside the room in question:
(7) [Krystal]: Would you look at me? Look at me, Adam. You've got your family locked outside this/that room. Now, I want you to open those doors.
In this case, contrary to (5) and (6), in order to refer to the room in which she is, Krystal no longer has the choice between this and that. Only the proximal demonstrative is possible. If the distal demonstrative is used, the referent changes and is necessarily some other room, most plausibly a case of anaphoric reference to a previously mentioned room (though a deictic interpretation would not be impossible if, for instance, the room in question was visible through a window from the room in which she is located). Contrary to what was the case in (5) and (6), where the speaker was outside the referent, no psychological/subjective factors can come into play, leading to a preference for that. Switching to that necessarily leads to a change in the intended referent.
Example (7) involves a demonstrative determiner, but the facts are the same when the demonstrative is used as a pronoun, as in (8a) or (8b):
(8)
(a) This is a beautiful room.
(b) That is a beautiful room.
If the speaker is outside the room, and even if she is just at the entrance of the room but has not yet stepped in, both variants of (8) can be used to refer to it, the choice being influenced by psychological/subjective factors. But if the speaker is in the room, no matter what psychological/subjective factors are at play, only (8a) is possible. Using (8b) would inevitably entail reference to a different room.
Thus, physical proximity and physical containment make very different predictions with respect to the choice of demonstratives. As previously mentioned, proximity of the referent to the speaker does not force a categorical choice. It is easy to find examples in corpora where a referent that is very close to the speaker is referred to using that. For instance, if the speaker has a book in each hand, she can say, waving first one of the books and then the other:
(9) Can you take this book back to the library and put that book in my office?
though obviously it is not the case that one of the two books is closer than the other. Similarly for pronominal uses:
(10) [Lily:] Yes, this is Charlie, and that is Mattie. [Jill:] (Chuckles) [Colin:] (Sighs) Oh, wow. Look at them. [Lily:] Yes. [Jill:] Now I may be biased, okay? But are these not the most precious babies you've ever seen? [Lily:] Yes, they are. (Young and Restless, 2011-01-04, SOAP)
The use of proximal and distal demonstratives in (10) is perfectly felicitous, even if Charlie and Mattie are both equidistantly close to the speaker. Both (9) and (10) involve a contrastive pair (as in footnote 5 above). However, that is possible independently of this contrast: in a situation that was identical except for the presence of a single referent, either the distal or the proximal demonstrative could have been used to refer to that referent.
Containment of the speaker in the referent, on the other hand, forces the choice of this, despite the fact that it is not necessarily clear that the speaker is closer to the referent when they are inside it than when they are just outside it.
We have illustrated this with the noun room in the previous examples, but any other noun denoting a physical space that can contain the speaker exhibits the same properties, e.g. in this country, in this town, in this house, in this garden, in this chair, etc. Note that physical containment must be understood in a broad sense: a person sitting in a chair is not literally inside the chair, but ordinary language ontology tells us that they are ‘in the chair’.
Previous discussions do not seem to have noted this difference in predictions made by proximity and containment. To take just one typical example, Fillmore (Reference Fillmore1997: 64), states that ‘For words that can refer to areas or spaces, the word “this” followed by the appropriate noun locates an object as being in the same area as the speaker is at coding time.’ This formulation is limited to determiner uses of the demonstrative and does not clearly distinguish proximity to the referent and containment in the referent, both of which can be characterized as ‘being in the same area as the speaker’.
Recall that the above discussion concerns deictic uses, which are usually taken to be the central use of demonstratives (see, e.g. Diessel Reference Diessel1999: 109ff. for discussion). In order to prevent possible misunderstandings, certain more marginal cases must be examined, where the boundaries between the deictic, anaphoric and recognitional uses are less clear. First, there are cases where reference is both deictic and anaphoric at the same time, for instance if a situationally accessible referent has already been mentioned in the previous discourse context, as was the case in (5) above, where the referent is referred to initially using the definite NP the doors and is referred to a second time using the demonstrative NP those doors, which is both anaphoric and deictic. It seems clear, however, that the immediate situational accessibility of the referent makes the deictic mode of reference dominant when the demonstrative NP is uttered. Consider now the following example, where A and B have just entered a room which they have not previously discussed:
(11)
-
A: I like this room. [Deictic]
-
B: Well, it's not so great, but I still think this room is better than the other one.
-
B's demonstrative NP might similarly seem to be both deictic and anaphoric. However, in this case, with spatial containment of the speaker in the referent, only proximal this is possible, suggesting that in fact the anaphoric interpretation is simply not available and the previous mention of the same referent is irrelevant.Footnote 7 B could not use that here for psychological/subjective reasons, e.g. with the intention of associating the referent with the addressee A who has a positive view of it. If, on the other hand, the speaker is in a room but is talking to the addressee on the phone, then deictic reference to the room is impossible, because the addressee does not have direct access to it. The recognitional and anaphoric uses, on the other hand, would be possible, as in (12a) and (12b) respectively:
(12)
(a) B: I'm in that room you talked to me about last week. [Recognitional use]
(b) A: Kim told me about a hidden room in the house last week.
B: Well, this is going to surprise you, but I'm in that room now. [Anaphoric use]
In cases like these, the speaker B can use the distal demonstrative to refer to the entity in which she is contained at the time of utterance because the mode of reference is not deictic.
Let us briefly consider, now, the case of the plural forms these and those. The following examples can be contrasted:
(13) [Cop:] Looks like no one's home in this one. [Detective Dallas:] Then use the passkey we got from the manager. There's a good chance our perp is hiding out in one of these rooms. (As the World Turns, 2007-05-10, SOAP)
(14) [Belle:] Shawn hasn't deserted us. He just – you know, he needed some air. [Duck:] Yeah, I know these rooms can get a little stuffy. That's why we put in all those windows. (Days of Our Lives, 2007-03-23, SOAP)
In (13) the detective and cop are at a motel or appartment building and can be imagined to be in a hallway rather than in any of the individual rooms. In this case, it would be possible to replace these by those. In (14), the protagonists appear to be in one of the rooms referred to by these rooms. In this case, it would not be possible to replace these by those without changing the reference to a set of rooms not including the one in which they are. It thus seems that the constraint applicable to the plural is an immediate extension of what we have seen for the singular: if the speaker is contained inside a member of the set of referents denoted by a plural demonstrative noun phrase, only proximal these is possible.
It can also be noted that the condition on physical containment of the speaker in the referent extends to the deictic use of the proximal demonstrative adverb here as opposed to distal there. If the speaker is contained in the referent at time of speech, only here is possible. At the entrance of a room, a speaker can refer to it deictically using here or there (Let's go in here/there) but once the speaker is in the room, only here is possible to establish deictic reference (I like it in here/#there). The demonstrative adverbs here and there appear not to have been studied in as much detail as this and that and would warrant further research.
3 Situational containment
As we have just seen, in the case of first-order referents, the idea of speaker containment within the referent applies transparently. In this section we turn to second-order referents, events and states, which we will refer to collectively as ‘situations’ (see Barwise & Perry Reference Barwise and Perry1983; also sometimes referred to as ‘eventualities’, see Mourelatos Reference Mourelatos1978). Because these are not first-order physical referents, simple spatial containment is not applicable to them. However, it appears that if the speaker is a participant in a situation, reference to that situation requires the proximal demonstrative this, suggesting that participants in events and states are conceptualized in ordinary language ontology as contained in the event or state. In some cases this notion of containment is more specifically conceptualized in mereological terms as being a part of, as in examples like I am a part of this investigation, I have every right to be a part of this discussion, She is a part of this conversation.
Consider the following example:
(15) [Brad:] (Chuckles) Jack, your political maneuvering already cost me half the value of those shares. And I suspect whatever offer you would make would be far below their actual value. [Jack:] Well, you know, they do say, ‘buy low, sell high.’ [Brad:] Not interested. [Jack:] Well, then I don't know who you're gonna find to take those shares off your hands. Dad can't afford it. Jill certainly can't afford it. Ash can't. No, the only one who is interested who can afford to offer you anywhere near market value is me. (Cell-phone-rings) (Ring) [Brad:] Hang on a second. You know, as much as I'm enjoying this conversation, Jack, I have to take this call. This may be somebody who's willing to make me a reasonable offer for the shares. (Young and Restless, 2005-09-22, SOAP)
The proximal demonstrative this with the deverbal noun conversation is used here to denote the event of conversing that is ongoing at the moment of utterance and in which the speaker is a participant. It would be impossible here to switch to that while keeping the same deictic mode of reference. Such a switch would necessarily lead to an anaphoric or recognitional use, which is very implausible in the context.
On the other hand, if the speaker was a third party listening to the conversation being referred to, without being a participant in it, she could have used either the proximal or the distal demonstrative to refer to it. Imagine that Sue and Ellen are witnessing Brad and Jack's conversation, it would clearly be possible for Sue to refer to it, addressing Ellen, using either the proximal or the distal demonstrative:
(16) [Sue:] I'm enjoying this/that conversation.
Here the choice between proximal and distal would be made depending on the usual psychological/subjective factors that are relevant in contexts not involving containment of the speaker.
By contrast, it should be noted that in (15), being a participant in the event preempts these usual psychological/subjective factors and makes it impossible to indicate any form of metaphorical closeness or distance to the referent through the alternation between this and that. Consider the following example:
(17) [Jen:] That's not what I'm doing. I'm letting him off the hook for trying to steal from you. He had your key! He stole it! [Lindsay:] I thought I lost that. Are you sure? Are you sure he stole it? Are you sure he was stealing [Jen:] Mom, check the security tape if you don't believe me. Rex is probably the one that put him up to it. [Lindsay:] No, that I don't want to believe. [Jen:] Why? Why? Because you still care about him? [Lindsay:] You know what? I'm not enjoying this conversation. In fact, I'm not even enjoying thinking about this conversation, so let's talk about doing something else. (One Life to Live 2004-10-05, SOAP)
In (17), the speaker is once again, at the moment of speech, a participant in the event being referred to. In other contexts, one might imagine that the negative stance taken towards the referent by the speaker (not enjoying) might lead her to prefer the distal demonstrative, metaphorically locating the referent far from herself. This kind of subjective distancing is not possible here though. The situational containment of the speaker in the referent forces the use of this and preempts psychological/subjective factors from playing any role in the choice.
The examples we have given up to now involve demonstrative determiners. Demonstrative pronouns exhibit the same properties, as illustrated in the following example:
(18) [Nikolas:] Who the – hey, get – get these people out – get – get the hell away from her! [Elizabeth:] Nikolas, they have to do this. [Nikolas:] How could you leave these people alone with her? You're not taking her away from me, you understand me? [Det. Harper:] Mr. Cassadine, I'm very sorry for your loss, but this is a murder investigation. (General Hospital, 2007-11-20)
Here, the speaker, Detective Harper, is the investigator in the investigation event, ongoing at the time of utterance, which is referred to by the deictic use of this. Once again, it would be impossible to replace this by that while keeping the reference constant.
In many cases, as seen with conversation in (15) and (17), and with investigation in (18), the event is referred to using a deverbal noun (other examples include this operation, this discussion, this inquiry, this trial, this action, this competition, this show, this race etc.). But there are also simple nouns that can denote events and exhibit the same behavior:
(19) [Jake:] […] Excellent. So can we just take a deep breath and relax and enjoy this meal? [Amanda:] Absolutely. (All My Children, 2010-03-31, SOAP)
In (19), where the meal is ongoing at time of speech and the speaker is a participant in it, only this is possible to establish the intended reference. The same is true for the game referred to in (20):
(20) [David:] I'm sure that your positive attitude will carry you a long way. I'm just not sure that it'll take you exactly where you want to go. Now whose turn is it? [Greenlee:] This game is boring. Besides, I think you're cheating. [David:] Sore loser. Ok, fine. I'll deal another hand. (All My Children, 2010-01-14, SOAP)
Similarly, the speaker can be contained in meteorological events, requiring the use of this (in this storm, in this rain, …).
The proposed analysis is further supported by constraints on exophoric uses of the VP anaphors do this and do that.Footnote 8 Consider the following example:
(21) – Hop on. – What? – Hop on. Now do this. – This? – Right. Arms around here. (COCA, MOV, 1990)
It would be impossible to replace do this in (21) by do that because the speaker is a participant (the agent) in the denoted event and is hence contained in it at the time of utterance, under the relevant conceptualization. Notice also that when the second speaker asks This? they are referring to the action that they are demonstrating in an attempt to satisfy the first speaker's request.
On the other hand, as expected, if the speaker is not a participant in the event, then once again exophoric reference allows a choice between this and that:
(22) Marlena: John, please stop doing that. It's really distracting. John: That's the point. (Days of Our Lives, 2008-02-07, SOAP)
In (22), that is clearly the unmarked choice, since the event is being performed by the addressee. Using this is not impossible, however, to indicate some type of speaker involvement in the referent; e.g. if the addressee was doing something to the speaker; see Gundel, Hedberg & Zacharski (Reference Gundel, Hedberg and Zacharski2004: 279).
Participation in stative situations exhibits similar constraints to those just discussed for events. Consider the following example:
(23) [Mimi:] Okay, I think I've got it. Please tell me that's the knife. [Max:] You have the pocketknife. Now get it out of the pocket. […] Okay, now, can you get the blade out? You're incredible. [Mimi:] Oh, yeah, you probably say that to every woman who sticks her hand in your pocket. [Max:] Hey, okay, now, less humor and more action, woman. I want to get out of this position. [Mimi:] It's hard. [Max:] Get it. Get it. Get it. […] [Mimi:] Oh, my God. I got it! Sit up! Sit up! [Max:] Okay. Mimi: Wow. Crazy, us being tied up like that, huh? (Days of Our Lives, 2006-12-19, SOAP)
As is clear from the example, the speaker is in the (obviously uncomfortable) position being referred to at the time of utterance. As expected, it would be impossible to replace this by that while keeping a deictic reference to the same stative situation. On the other hand, if it is the addressee who is in the position at the time of utterance, both this and that are possible and the choice will be based on psychological/subjective grounds, e.g. using this might mark a form of empathy with the addressee:
(24) I want you to get out of this/that position.
In our discussion of (12) above, where speaker and addressee were talking over the phone, we noted that there are contexts where a referent is immediately accessible to the speaker but not to the addressee. Reference to stative situations exacerbates these issues because they are typically less accessible to direct perception in a given situational context than first-order entities or events.Footnote 9
This means that in many cases where the speaker is in a given state, the addressee will only have access to the precise nature of that state through an anaphoric mode of reference. This problem makes it necessary to be especially careful in examining demonstrative NPs denoting states. Consider the following example:
(25) [Angie:] Look, I just want to help you. Do you know what they're probably talking about right now? [Krystal:] Well, I don't know, but, I mean, she might be telling David that he's the father of that baby. [Angie:] Uh-huh. And if that's the case, and the child survives, maybe finally you'll be free of him. He'll have what he wants. He can stop fixating on Babe's child and you. [Krystal:] If you think that I am trying to get out of this situation, then you just do not understand. I love David. [Angie:] Then why did you come? [Krystal:] To find out the truth. (All My Children, 2009-03-04, SOAP)
As is clear from the example, the speaker is in the situation being referred to at the time of utterance. But this is not what makes the situation accessible to the addressee. Rather it is the previous discourse context that makes it clear exactly what situation is being referred to. Thus the reference established by proximal this here is anaphoric rather than deictic and, as a consequence, despite the fact that the speaker is contained in the referent state at time of utterance, a choice is possible between this and that based on psychological/subjective factors, so that the variant with that would also be possible:
(26) If you think that I am trying to get out of that situation, then you just do not understand.
The difference between (23) and (25) is that in the former the state being referred to is a physical state of the speaker that is immediately accessible to the adressee. This makes deictic reference not only possible in (23), but necessary, preempting an anaphoric mode of reference, and making the proximal demonstrative this the only one possible. In (25), on the other hand, the mode of reference can be construed as anaphoric, making both proximal and distal acceptable.
We can further illustrate the conceptualization of stative situations as containing the speaker thanks to linguistic expressions of the type ‘X be in state’ or ‘X get out of state’, e.g. get out of this mess, get out of this trap, get out of this profession, get out of this marriage, get out of this deal, get out of this relationship, etc. Similarly for various psychological states: get out of this mood, get out of this rut, this happiness is too much to bear, …
It appears then that being a participant in a situation (event or state) at time of utterance is construed in linguistic ontology as being contained in that situation and forces the use of the proximal demonstrative this to refer to it deictically. If the speaker is not a participant in the situation, on the other hand, in deictic uses, there is a choice between distal and proximal which can be made in terms of proximity (physical or psychological/subjective). Anaphoric uses allow the same choice between proximal and distal since the referent is not accessed on the basis of its containment of the speaker at time of utterance. In this context, one might wonder if the relevant notion of containment extends to third-order referents, e.g. facts or propositions. This does not appear to be the case, simply because such referents do not seem to allow deictic uses at all. Expressions like this fact or that fact can only be used anaphorically since their referents are necessarily linguistically constructed.
4 Set containment
We now turn to the case where the demonstrative NP denotes a set of humans. In such cases, the speaker either is or is not a member of the set at the time of utterance. Consider (27):
(27) Because in spite of all the things I've done wrong, everything I've been through, the one thing that has meant the most to me, that's kept me going, is being a part of this family, is being a Martin. (All My Children, 2011-03-02, SOAP)
In this example, the speaker is a member of the family (this individual-level property obviously holds at time of utterance) and is as such contained in the referent. Because this is a clear case of exophoric reference (the addressees, who are members of the same family, have no problem accessing the referent situationally), only the proximal demonstrative is possible. Replacing this by that in (27) would necessarily lead either to exophoric reference to a family that is not that of the speaker or to anaphoric reference to the speaker's own family in a situation where the referent is not immediately accessible to the addressee, neither of which has any plausibility in the context. Other nouns that exhibit similar behavior include crowd (I had trouble finding you in this crowd), company, team, household, community, administration, class, marriage, relationship, …Footnote 10 In some instances, set membership is expressed in what appear to be mereological terms, e.g. You are a big part of this company. In all of these cases, then, speaker containment is understood as membership of the speaker in the set referred to by the demonstrative NP and forces the choice of proximal this in deictic uses.
5 Temporal containment
Huddleston & Pullum et al. (Reference Huddleston and Pullum2002: 1559–62) provide a thorough discussion of the use of temporal nouns. As they point out, when the intended referent is a time interval, if the speaker (or more precisely, the time of utterance) is contained within the referent, only this is possible (see also Fillmore Reference Fillmore1997: 68ff.):Footnote 11
(28)
(a) I ate lunch at Harvey's twice this week.
(b) I ate lunch at Harvey's twice that week.
In (28a) reference is deictic to the ongoing week at the time of utterance. In (28b), on the other hand, reference can only be understood as anaphoric to a previously mentioned week.
The situation is thus parallel to what was discussed above for physical referents, except that containment is temporal. On the other hand, when the speaker is not contained in the time interval being referred to, there is a specific subsystem. First, as just illustrated for (28b), that is never possible in deictic uses.Footnote 12 Rather, a demonstrative NP like that week can only be used anaphorically (29a) or in shared knowledge contexts (29b).
(29)
(a) We spent a week in Spain. Due to the covid outbreak, that week turned out to be a nightmare.
(b) [Seeing a colleague for the first time in a month, when your last interaction was a truly terrible week at work] I'm so glad to see you looking so good. That week was really something. You were completely exhausted at the end of it.
Second, for deictic reference to time intervals other than those including the time of speech, a set of specific expressions is used, namely, last week (= the one before the one containing the time of speech), next week or this coming week (= the one after the one containing the time of speech), the week before last, etc.Footnote 13
It appears then that the use of proximal this is forced by temporal containment of the time of utterance in the referent and, by extension, of the speaker at time of utterance, in parallel to what was previously shown for spatial containment.
6 Conclusion
In this article, I have brought to light a hitherto unnoticed constraint on the deictic use of demonstratives in English that forces a categorical choice of the proximal demonstrative this, namely those cases where the speaker is contained in the referent at time of utterance. Containment can be understood as spatial, situational, temporal or set membership.
In their recent article on demonstrative reference in written registers, Maes, Krahmer & Peeters (Reference Maes, Krahmer and Peeters2022b) point out that there are a series of contexts in which only proximal this is possible with endophoric demonstratives. Two of the cases they discuss are clearly related to what has been presented here. First, they note (p. 200) the case of what they call ‘origo demonstratives’, occurrences like this week, this era, this room or this country (though they provide no specific explanation as to why only the proximal variant is possible). I would consider such cases as simply exophoric, though this is only a matter of terminological preference. In any case, they are clearly explained in terms of the proposals made in sections 2 and 5 above. Second, they discuss (p. 201) the case of what they call ‘self-reference demonstratives’, occurrences like this chapter, this article, this manual etc.Footnote 14 Such occurrences are clearly linked to textual registers and do not occur in the spoken dialogue register investigated in this article. However, it seems that the analysis proposed here immediately extends to these cases. Specifically, I would argue that this is an extension of spatial containment of the speaker. Just as the speaker is metaphorically at a certain place in the time line, which can be included in temporal referents (this week, this month), the speaker is metaphorically at a certain place in the text, which can be included in textual referents (this chapter, this section). Further justification for the parallelism with temporal referents comes from the fact that similar expressions can be used to refer to a textual referent that does not include the place where the speaker is in the text, but rather is adjacent to it, in particular previous and next (in the previous chapter, in the next chapter).Footnote 15
To conclude, as previously mentioned, traditional analyses of the proximal/distal alternation for English demonstratives have been based on the distance (physical or psychological/subjective) between speaker and referent. Though this is undoubtedly the crucial factor at stake in most occurrences of demonstratives, it does not provide a categorical criterion for the choice, especially with referents that are close to the speaker. The property of containment, on the other hand, when it is applicable, does provide a categorical criterion of choice, forcing the choice of this in deictic uses when the speaker is contained in the referent at time of utterance. The relevant notion of containment can be either spatial, in which case the speaker is physically inside the referent, or figurative. For the latter case, three subcases have been shown to be relevant, namely, situational containment, where the speaker is a participant in the denoted event or state; set containment, where the speaker is a member of the denoted set; and temporal containment, where the speaker (or, more precisely, the time of speech) is included in the denoted time segment.
Most studies have assumed that the physical interpretation of the proximal/distal distinction is basic and that other interpretations are metaphorical extensions. Cheshire (Reference Cheshire1996) and Cornish (Reference Cornish2001), on the other hand, claim that the ‘modal’ sense (Cornish's term) is basic and the proximal/distal physical sense is derived.Footnote 16 The present study provides a further piece of evidence that the locative use is basic since in those situations of speaker containment where the choice of the proximal demonstrative is the only possibility, it actually preempts any ‘modal’ (or ‘psychological/subjective’) stance based on the choice between this and that. In deictic uses of demonstratives, modal effects of metaphorical speaker closeness are only possible when the speaker is not contained in the referent at time of utterance.
