TEMPORAL CONSTRAINTS IN THE USE OF DEMONSTRATIVES IN IBERIAN SPANISH *

Natural linguistic data show that a close relationship exists among demonstratives and time in Spanish. Such connection appears to be of a pragmatic nature whereby linguistic expressions that encode or denote time license the use of demonstratives under certain conditions. Generally speaking, these linguistic expressions or licensors contribute to setting up adequate temporal frames of reference for a felicitous use of demonstratives. Although grammatical tense appears to be the principal type of licensing expression other linguistic elements (i.e. adverbials, noun phrases, etc.) also play an important role in establishing the reference time of the discourse. The aforementioned connection time-demonstration occurs when speakers employ demonstratives to refer to objects within the textual world (i.e. discourse referents). In this paper I explore the temporal constraints that impose restrictions in the use of demonstratives in Spanish and argue in favor of characterizing the relationship timedemonstration as a discourse deixis phenomenon in close connection with time deixis. The co-referential link that gets established between the demonstrative and the temporal expression is formalized and a presuppositional characterization is proposed for the three demonstratives in their discourse-deictic use.


TEMPORAL CONSTRAINTS IN THE USE OF DEMONSTRATIVES IN IBERIAN SPANISH *
Iker Zulaica-Hernández Indiana University-Purdue University, Indianapolis ABSTRACT.Natural linguistic data show that a close relationship exists among demonstratives and time in Spanish.Such connection appears to be of a pragmatic nature whereby linguistic expressions that encode or denote time license the use of demonstratives under certain conditions.Generally speaking, these linguistic expressions or licensors contribute to setting up adequate temporal frames of reference for a felicitous use of demonstratives.Although grammatical tense appears to be the principal type of licensing expression other linguistic elements (i.e.adverbials, noun phrases, etc.) also play an important role in establishing the reference time of the discourse.The aforementioned connection time-demonstration occurs when speakers employ demonstratives to refer to objects within the textual world (i.e.discourse referents).In this paper I explore the temporal constraints that impose restrictions in the use of demonstratives in Spanish and argue in favor of characterizing the relationship timedemonstration as a discourse deixis phenomenon in close connection with time deixis.The co-referential link that gets established between the demonstrative and the temporal expression is formalized and a presuppositional characterization is proposed for the three demonstratives in their discourse-deictic use.

Introduction
In Iberian Spanish, I'll argue, a pragmatic connection occurs when the Spanish distal demonstrative determiner and pronoun aquel (pron.that yonder, det.that NP yonder) find expressions that encode a past time like grammatical tense markers, temporal adverbials that describe a position in the time line or, generally speaking, expressions encoding time in the nearby context of the demonstrative.These timeencoding expressions appear to become the necessary 'licensor' for a felicitous use of the demonstrative in discourse and serve to configure contexts so they may play the role of temporal frames of reference for a pragmatically felicitous use of the distal demonstrative.
Crucially, what all the aforementioned licensing expressions have in common is that they all semantically encode a past time thus confirming that the observed interaction may transcend grammatical tense and should be more precisely defined as a connection between a past time and the distal demonstrative anaphor. 1 The use of the other Spanish demonstratives, proximate este 'this' and medial ese 'that', are not constrained by this licensing condition so the presence of a past time encoding expression in the text is just a necessary condition for the use of the distal demonstrative.Consider the short dialogue in (1): (1) [An interview with an Opera singer] Q: ¿Disfruta usted más en los ensayos que en las representaciones en público?'Do you enjoy rehearsals more than performing in front of the audience?'A: Me encantan los ensayos, es cierto.Amo este/ese trabajo.(#aquel)'I love rehearsals.It's true.I love that job.' [Corpus del Español: Interview, ABC journal] Notice that the distal demonstrative aquel 'that yonder' can't be used felicitously in a discourse that is clearly referring to a situation or event that has current temporal relevance (i.e. the interviewee refers to her present job.)Now, if we change the temporal frame of this brief discourse by changing the tense of the verbs to obtain (2) the distal demonstrative is fine.Use of the medial demonstrative ese 'that' is also felicitous showing that it is not constrained by such temporal parameter.Use of the proximate este 'this' shows some variability regarding the speakers' judgment about its felicitous use. (2) [An interview with an Opera singer] Q: ¿Disfrutaba usted más en los ensayos que en las representaciones en público?
1 The alleged correlation 'tense-demonstratives' does not appear to be restricted to Spanish.Kirsner et al. (1987) studied the effect than tense may have on the speaker's choice of demonstratives deze (this) and die (that) in Dutch.At the intrasentential level, they found that 59% of present tense verb forms cooccur with proximal deze and 67% of past tense forms co-occur with distal die.These figures are based on the scrutiny of 43-deze sentences and 42 die-sentences containing non-perfect verb forms.Nevertheless, as the authors of the study point out, when context is added, other factors such as referential distance or the degree of detail with which the referent has been described override the influence of tense on demonstrative choice.It is worth noting that for these authors the differences in meaning between Dutch demonstratives can be fundamentally explained on the basis of "the degree of attention which the addressee is instructed to give to the referent of the noun" (1987: 17).Their paper poses interesting questions regarding the possibility that other factors may influence the correlation tense-demonstration.
demonstrative and past time denoting expressions within the framework of Temporal Discourse Models.Finally, in § 6, I provide a presuppositional characterization for Spanish demonstratives based on natural data from the corpus and argue in favor of considering these elements as anaphoric presupposition triggers.

Spanish demonstratives, time and context
Demonstrative expressions and natural language tenses have a deictic nature.Both linguistic elements need to be interpreted relative to other element(s) of the context of utterance.The content of demonstratives is usually characterized in terms of the higher or lesser degree of spatial proximity of the speaker with respect to the object pointed at in the utterance situation or the demonstratum.This constitutes perhaps the most typical use of demonstratives in which one the speaker accompanies her utterance of the demonstrative with an ostension such as a nod of her head, or a pointing with her index finger to demonstrate some specific (commonly discrete, three-dimensional) object to the addressee.3Thus, for example, the English binary system of demonstrative determiners this-NP/that-NP is intended to mark the differential proximity within the context of utterance of the object pointed at with respect to the speaker (the deictic center).But while some languages instantiate binary demonstrative systems others have tripartite systems.This is the case of Spanish, which instantiates a more complex system of three terms este-NP, ese-NP and aquel-NP.Traditionally, Spanish demonstratives have been characterized in terms of the different valuation of the features [+ speaker]/[+ addressee] and [± proximity].The system is illustrated in Tables 1 and 2. Denotes proximity of the object with respect to the second person (the addressee) That N Aquel [-Proximate] Denotes that the demonstrated object is distant with respect to both speaker and addressee

That N yonder
The notion of time is linguistically expressed in different forms cross-linguistically (Klein 1994).Grammatical tense markers, time adverbials or lexical aspect are only a 199 few of the different linguistic mechanisms used to express time in natural language.At the discourse level, it is widely accepted that temporal structure greatly contributes to discourse construction in different ways.Thus, for example, the sequence of clauses in narrative speech tends to match the way in which the events actually occurred, which contributes to overall discourse coherence and cohesion.See the work by Asher & Lascarides (2003), Asher, Prévot & Vieu (2007) or Mani & Pustejovsky (2004) for different views on how to account for temporal relations and discourse coherence.On the other hand, temporal information is also crucial to interpret a variety of linguistic elements in the unfolding discourse.This is clearly the case of deictic elements, which rely on a contextual parameter for their interpretation (i.e. the words 'now' or 'tomorrow' cannot be interpreted without reference to some specific time.)In deixis proper, the parameter that is needed for interpretation is most commonly found in some extra-linguistic element of the physical context (i.e. the interlocutors, spatial coordinates, etc.) but other expressions, of an anaphoric nature, need a suitable time-denoting antecedent in the preceding text to be resolved.See, for example, the seminal work on temporal anaphora by Partee (1973Partee ( , 1984)), Webber's (1988) ideas on tense as discourse anaphor or the study on temporal locators by Alves (2006), inter alia.Whether deictic or anaphoric in nature what all these expressions have in common is their dependency on contextual information to be fully interpreted in the discourse.

The relationship tense-demonstratives
The Spanish grammarian Fernández-Ramírez (1951) was the first to observe and describe the apparent relationship that arises between certain demonstrative expressions (i.e.demonstrative determiners and pronouns) and some specific tenses in Iberian Spanish written and spoken discourse.As Fernández-Ramírez noted, there appears to be a tendency whereby the proximal demonstrative combines better or more frequently with present tenses whereas the distal demonstrative strongly tends to combine with past tenses.The medial demonstrative appears to show a more balanced behavior and equally combines with present and past tenses.In summary, some specific combinations of demonstratives and tenses are strongly preferred while other combinations may seem unnatural to Spanish speakers. 4ative speaker intuitions seem to confirm the author's hypothesis and a careful analysis of corpus examples support it empirically.Our first observation of natural linguistic data indicates that the distal demonstrative aquel is a marked element in that it shows a strong preference to combine or co-appear with past tenses in the clause or discourse.On the other hand, the proximate and medial demonstratives este and ese seem to combine well with any kind of tense (i.e.present or past), although the proximate element tends to favor present tenses.Examples (3)-( 6) illustrate the felicitous combinations of proximal and medial demonstrative NPs with present and past tenses.Example (7) shows a combination of the distal demonstrative with a past tense, which constitutes the most natural combination for the distal demonstrative in modern Iberian Spanish.The specific demonstrative-tense combinations are marked in italics.
Para algunos dirigentes del partido este hecho convierte a la conferencia "en un mero acto de cosmética" e impide que se pueda profundizar en la discusión sobre el objetivo estratégico y programático del partido.'No debates or voting are planned for today's session.For some of the party leaders, this fact makes the conference "a mere act of cosmetics" and it prevents them to discuss in depth the party's programmatic and strategic target.As we expected, the preferred combinations of demonstratives and tense are not restricted to demonstrative determiners and an identical pattern can also be observed with demonstrative pronouns.Again, both proximate esto 'this' and medial eso 'that' combine felicitously with past and present tenses whereas the distal pronoun aquello 'that yonder' shows a clearly marked predilection to appear with past tenses in natural written and spoken discourse.This pattern is shown in examples ( 8)-( 12).
In (8), a series of past tenses indicate that the event occurred in the past at some time prior to the utterance time.In the last sentence, we observe that a present perfect tense is used and the demonstrative expression este evento refers back to the noun phrase La XVI edición del Camel Trophy.In this case, the proximate demonstrative is combined with a present tense in the last sentence, what indicates that Spanish speakers use the proximate and medial demonstratives to refer to past events even when the demonstrative does not specifically combine with a past tense in a sentence.When reference is made within the text (i.e.intratextual or endophoric reference) occurrences of distal aquel/aquello along with present tenses in natural discourse are scarce, and even when they can be found the whole discourse appears to be framed within the past.This is shown in ( 12), where the distal demonstrative noun phrase aquel hombre 'that man' appears in the last sentence along with a present tense hablamos 'we talk about' and with a present tense in the prior sentence as well surge 'appears'.That the discourse appears to be framed in the past it is indicated by the starting clause and around 230.000 and 150.000 years ago and the past tense estaban 'were'.The present tenses here and in many other cases seems to be a case of Historical Present, that is, the use of a present tense with a past temporal reference.
(12) Y hace aproximadamente entre 230.000 y 150.000 años, mientras que los herederos del Erectus en África estaban a punto de dar paso al Homo sapiens moderno, surge en Europa la especie que hoy por hoy más fascinación despierta y que tantas incógnitas presenta.Hablamos de aquel hombre poderoso, alto, rubio... ¡el hombre de Neandertal! 'And around 230.000 and 150.000 years ago, while the heirs of the Erectus in Africa were about to take the leap to modern Homo Sapiens, the species that are considered the most fascinating and mysterious species today appears in Europe.We talk about that powerful, tall, blond man ¡The Neanderthal man!' (Los misterios de la evolución humana, 2001) At least one element encoding a past time could be found in all the examples analyzed thus far.In many cases, more than one element encoding a past time coappeared in the anaphor's immediate context (i.e. a sequence of past tenses in a narrative discourse).The fact that all these examples feature some linguistic expression that either directly or indirectly denotes a past time contributes to positioning the event talked about at some point in the past of the speech time.Cases of the distal demonstrative aquel in Peninsular Spanish which do not feature at least one linguistic element that encodes a past time are very difficult to find in written and spoken discourse so this may be an indication that the distal demonstrative aquel is strongly favored in contexts where a past temporal frame has been established via past tenses, adverbial expressions (dates, time adverbials, etc.) or background knowledge.

Reference and temporal deixis
As referring expressions, adnominal and pronominal demonstratives this (NP) / that (NP) can be used in two basic modes that allow speakers to refer to entities in different discourse situations.In the deictic proper mode, speakers commonly use demonstratives to refer to physical, concrete entities in the real-world speech situation.Utterance of the demonstrative (very likely accompanied by a pointing gesture) has an important communicative effect, namely, that of focusing the attention of the addressee on a particular entity in the perceptual or visual field.This is accomplished by making the intended entity salient from among a set of (potentially) competing entities.Thus, both speaker and addressee focus their attention on the same element and the speaker's intended communicative goal is achieved.This is the socalled joint attention effect in the psycholinguistics literature (Diessel 2006).When employed exophorically, the role of the pointing gesture (a pointing finger, a gaze or movement of the head) may become essential.It completes the meaning of the demonstrative expression and serves to disambiguate the speaker's reference.Demonstratives can also be used exophorically without an accompanying ostension, but in such cases the entity referred to is already sufficiently salient in the visual field for the interlocutors to have focused their attention on it and, consequently, an accompanying gesture by the speaker would be redundant or irrelevant to achieve the intended communicative goal.
There are certain uses of demonstrative elements that depart from the purely deictic mode.These uses, long and widely recognized, have been characterized as anaphoric or discourse anaphoric/deictic by different authors (see, inter alia, Asher 1993;Diessel 1999;Janssen 1996;Gundel et al. 2001Gundel et al. , 2003;;Hegarty et al. 2001.)As demonstrative anaphors, demonstratives are coreferential with a range of textual elements.From a syntactic point of view, the antecedents of demonstrative anaphors can be of a varied nature: NPs, subordinate clauses, entire sentences or larger textual fragments.Semantically, these antecedents comprise a rich ontology that ranges from concrete entities to propositions, facts, events, event-types or even discourse topics. 5hether the anaphoric referring mode is derived from a primary deictic character or not is an issue quite beyond the purpose of this paper, but studies on language acquisition indicate that the deictic features are learned at the earlier stages (Diessel 2006).
The only difference between the referential capabilities of deictic and anaphoric uses of demonstratives lies in that these capabilities have been transferred from a realworld context of utterance common to strict deictic uses to a textual (endophoric) domain in the anaphoric use.In the two cases, the primary communicative function appears to remain the same and the pointing gesture, absent in demonstrative anaphors, may have evolved into derived pragmatic functions in the anaphoric use. 6s for Spanish demonstratives, Eguren (1999: 946) acknowledged the existence of anaphoric uses of Spanish demonstratives that depart from strict deictic ones.Nevertheless, the dual deictic/anaphoric character of demonstratives is still today a source of much debate among linguists and a generally accepted distinction for both uses is still far from being developed.Janssen (1996: 80) provides a basic characterization of the distinction between deictic and anaphoric uses of indexical elements as can be found in most literature on the topic today.As the author points out: Deictic use of an indexical element: An indexical element is used deictically if it is referentially related to information derivable from the current situational frame of reference, which is cognitively accessible to a basically constant set of speaker(s) and addressee(s).

Anaphoric use of an indexical element:
An indexical element is used anaphorically if it is referentially related to information derivable from the current text-based frame of reference, which is cognitively accessible to a basically constant set of speaker(s) and addressee(s).
Note that the fundamental distinction between the two modes lie in the different frame of reference for a particular expression, namely, the current situation in the case of deixis and the text in the case of anaphora.Diessel (1999) distinguished between anaphoric and discourse deictic uses of demonstratives.For him, three main distinctive features characterize discourse deictic demonstratives, namely, they make reference to propositions or speech acts; they link two propositions, and their referents do not commonly persist in discourse.Himmelmann (1996) studied several unrelated languages and established a universal typology of demonstrative uses, namely, situational (purely deictic), non-situational (anaphoric, cataphoric, discourse deixis and a new recognitional type proposed by the author whereby the intended referent of the demonstrative is to be identified via specific, shared knowledge rather than through situational clues or preceding discourse segments).
Some of these derived uses of demonstrative elements contribute to discourse cohesion in various ways.Thus, demonstrative determiners and pronouns enter in coreferential chains with varied discourse entities.Consider example (13), where the first sentence is the antecedent (the intended referent being the event denoted by the first sentence) of the demonstrative pronoun aquello in the second sentence.The subscript indicates coreferentiality between antecedent and anaphor.7 (13) [Mi abuela pasó como siete días muriéndose] k .Aquello k fue horrible.
'My grandmother took seven days to die.That was horrible.'(Entrevista 101.Venezuela, Oral) It is common practice in grammar studies to differentiate the semantic content of the different deictic demonstratives by virtue of a ± proximity value.Thus, for example, the English binary system of demonstratives is characterized as having one [+proximate] demonstrative determiner this-NP and the [-proximate] that-NP.As we have shown in Table 2, Bello (1892) characterized the tripartite system of Spanish demonstratives along the same lines.The proximity condition would be explained as the ± spatial/temporal distance of the demonstrated object (the demonstratum) relative to the position of the speaker in the utterance situation.In most cases, the spatial magnitude is characterized as the speaker's perspective in terms of physical distance with respect to the object pointed at.But the notion of distance does not have to be restricted to the spatial domain or be taken as an absolute value.Thus, for example, the degree of distance conveyed by the use of the proximate demonstrative este 'this' with respect to the speaker is to be better conceived as a relative «region of proximity» that may vary according to the expression used: esta casa 'this house', esta ciudad 'this city' or even este mundo 'this world' (Eguren 1999: 940).Likewise, the varying degree of distance can also be conceived of as temporal remoteness or closeness to the speaker; the so-called deixis am phantasma (Bühler 1990).This type of deixis is illustrated in ( 14).Here, by using the demonstrative expression aquellos tiempos 'those times', the speaker appears to be referring to some specific point in the past (or more specifically, to some situation or event temporally located in time t where t < utterance time.)(14) Afortunadamente aquellos tiempos trágicos ya pasaron para siempre.
'Fortunately, those tragic times had passed forever more.' (La Vanguardia, 22/11/1994) I'll argue that the particular use of the distal demonstrative aquel in Spanish may be described as a particular case of temporal deixis (Fillmore 1975).As we have seen so far, the distal demonstrative in Spanish appears to be a term whose primary function in the discourse is to make reference to a past time in some sort of temporal deictic use.The distal demonstrative would accomplish its temporal deictic use indirectly by referring to a discourse entity which is anchored at some unspecific past time (prior to the utterance time) or some specific time if such a time is introduced in the discourse via some specific time denoting expression (an adverbial expression, a calendar date, etc.)However, the ability to refer to discourse referents that are temporally located in the past is not restricted to the distal demonstrative.As we have seen in previous examples, the proximate and medial demonstratives can also be used for this specific referential purpose.What makes distal aquel/aquello exceptional as a referring expression is that it appears to be very much restricted to this particular use in modern Iberian Spanish.

Corpus study: data and analysis
The goal of the corpus study is to assess whether the particular behavior of the distal demonstrative that I have presented in this paper can be supported empirically by exploring the relationship between the distal demonstrative and tense in discourse.The results of this corpus study are not intended to be conclusive but only a first approximation to the use of the distal demonstrative in modern Iberian Spanish with an aim to unveil the mechanisms that impose restrictions on its use in discourse.The study presented here has limitations due to the nature of the corpus itself and to the possibility that other factors besides tense may have an effect in the way Spanish speakers use the distal demonstrative.Therefore, other empirical studies will need to be done to characterize other potential restrictions in the use of the distal demonstratives aquel/aquello.
The corpus CREA (Corpus de Referencia del Español Actual) has been the source of Spanish data used throughout this paper and for all our corpus samples and illustrative examples.The CREA corpus of Spanish is a very large collection of texts.A dedicated search interface allows the user to search the corpus for words and phrases and display the search result as a concordance with limited context, but the sufficient amount of context for the purposes of this paper.The corpus comprises a variety of written texts (newspaper, novels, emails, etc.) as well as transcribed spoken discourse (interviews, speeches, etc.).For the purposes of this paper, we have included both written and spoken discourse in our corpus samples.This is the most comprehensive corpus on modern Spanish that can be consulted today in terms of size and scope of variation, although it also shows limitations due to its absolute lack of annotation.The corpus has over 160 million words.Around 50% of the materials in the corpus come from Iberian Spanish sources and 50% from Latin American Spanish sources.90% of the corpus corresponds to the written language and 10% to oral language.
The findings of our study pose interesting additional questions concerning the nature and current linguistic status of the distal demonstrative in peninsular Spanish.A quick, superficial analysis of any corpus of Spanish language would soon reveal that the distal demonstrative determiner aquel and its pronominal counterpart aquello are somehow marginal in terms of frequency of use.The corpus revealed the following overall figures regarding the frequency of occurrence of the three demonstrative pronouns and determiners in Iberian Spanish.TOTAL 229.853 1.0 217.531 1.0 12.322 1.0 The data presented in these tables 3 and 4 is limited to the Spanish spoken in Spain.The figures show the low frequency of use of the distal demonstrative determiner aquel and the demonstrative pronoun aquello when compared with the two other demonstratives.When European Spanish is left out of the picture the results from the rest of the Spanish-speaking world show similar results, as shown in Tables 5 and 6.8The second part of the study is limited to European Spanish in written and spoken discourse and was also limited to the three demonstrative pronouns: esto 'this', eso 'that' and aquello 'that yonder'.The overall number of demonstrative tokens in the corpus is too large so the only way to size down the sample to a manageable number was by selecting a small group of verbs.Two types of verbs have been used, namely, a group of verbs suceder 'happen', ocurrir 'occur', terminar 'finish', etc., which force a reading where the demonstrative refers back to a discourse entity that denotes a higher order entity (fact, event, accomplishment, etc.) possibly realized nominally or propositionally in the text, and a second group comprising the verb ser 'to be', which does not select for any specific arguments.Thus, arguments of the verb ser can be any 207 type of concrete or higher order entity.The purpose of choosing these two verb groups was to control for differences in the semantic denotation of the referring expression.The three demonstrative pronouns were combined with different verb tenses to come up with [demonstrative + tense] combinations like: esto sucedió 'this happened'; esto sucede 'this happens'; esto ha sucedido 'this has happened', etc.The verbs and tenses employed for this study are shown in table 7. The two tense groups were categorized according to a temporal deictic dimension.Thus, PRESENT tenses are those commonly characterized as totally or partially overlapping the utterance time in a linear, two-dimensional representation of time.Events denoted by present simple and present continuous tenses overlap the utterance time, whereas the present perfect tense -while technically a past tense-can be conceived of as proximate to or even overlapping the utterance time. 9On the other hand, the group of PAST tenses includes simple (aspectually divergent imperfect and preterite) and compound tenses.Semantically, the event arguments of all these past tenses lie at some time prior to the utterance time in a linear representation of time.Notice that two subjunctive tenses have been included in the present study, present and past (imperfect) subjunctive, but in an independent data set (see Table 9) as subjunctive tenses are not commonly characterized along a temporal dimension. 10They are temporally void tenses in some respect, and the events denoted by them belong in the realm of the hypothetical or non-factual information.In consequence, they cannot be clearly located in the past of or overlapping the utterance time.My having included these two tenses in the study can be explained for two reasons.On the one hand, Spanish grammar terminology still makes a distinction between a present of the subjunctive and a preterite of the subjunctive, hence implying some temporal distinction between them.On the other hand, the two subjunctive tenses 9 The Spanish pretérito perfecto compuesto is ambiguous between, at least, two possible readings, namely, one reading in which the past event described includes/extends into the utterance time (i.e. the event is said to have current relevance); and a second possible reading in which the event described lies at a recent past time (i.e.recent with respect to the utterance time but not overlapping or extending into the utterance time).We have not controlled for this subtle, however important, distinction in our corpus study and we will certainly include this variable in future extensions of this paper.However, we think that the overall results of the quantitative study, in particular the marked status of distal 'aquel', do not get generally affected since the low number of combinations of the distal demonstrative and the pretérito perfecto compuesto (i.e. the English present perfect) analyzed barely have an effect on the total figures.For a detailed account on the Spanish present perfect (Span.Pretérito perfecto compuesto or antepresente) see Alarcos-Llorach (1970), Moreno-Cabrera (1999), or Brugger (2001). 10The Spanish verb system makes a basic distinction between two modes: the indicative mode and the subjunctive mode.All indicative tenses convey the modal notion of factuality of the action or event denoted by the verb whereas the subjunctive tenses convey the modal notion of unreality (or nonfactuality) of the action or event denoted by the verb.Thus, for example, we find a present indicative and a present subjunctive in Spanish.
208 might be used contrastively to compare their results with indicative tenses.The observed frequencies are presented in Tables 8 and 9. Overall, it is worth mentioning the low number of occurrences of distal aquello found in the corpus specially when the figures of the two other demonstrative pronouns are thrown into the picture.This goes in line with Tables 3-6, where raw figures indicate that, generally speaking, the distal demonstrative is much less frequent that the proximate and distal demonstratives in Spanish written and spoken discourse.Only 655 combinations aquello + tense (4.5%) could be found out of 14.627 specific [demonstrative + tense] combinations in the corpus.Despite the small sample of tenses (n = 6), the sample (n = 14.627) is large enough to confirm that distal aquello is the less frequently used element of the tripartite system of demonstrative pronouns in Iberian Spanish.Interestingly, 91% of all cases of aquello, leaving aside the subjunctive, were found in combinations with past tenses.The total observed frequencies are shown in Table 10.The distribution in Table 10, which is highly significant (X 2 = 2845.767(df = 2), p < .0001),shows extremely few cases of combinations [aquello + past tense], which may be an indication that past tenses actually constraint the use of the demonstrative in discourse as it is argued in this paper.Other interesting conclusions can be drawn from the distribution in Table 10.The observed frequencies in Tables 3 and 5 show that the demonstrative pronoun eso 'that' is the most frequently used pronoun of the tripartite system in Spanish but our figures in Table 10 show almost identical total numbers for esto (n = 6.973) and eso (n = 6.999).But more interestingly, pronoun esto clearly disfavors to be combined with past tenses.Note the total number of combinations with past tense for esto (n = 624), eso (n = 1.310) and aquello (n = 597).This is particularly noticeable when we consider the high overall number of occurrences of esto compared with aquello.On the other hand, demonstrative eso, while strongly favors to be combined with present tenses, doubles the numbers of [esto + past tense] combinations.The relative frequencies of the three demonstratives combined with past and present tenses are shown in Table 11, and graphically in Figure 1.In summary, our frequencies show a strong preference for the distal demonstrative aquello to be combined with past tenses in discourse, which may support the idea that distal aquel/aquello is strongly favored or licensed in contexts that set up a past temporal frame of reference, like (past) narrative contexts or simply contexts that describe past events, facts or situations, possibly realized nominally or propositionally in the text.In these contexts, the demonstrative would be used discourse-deictically to refer to these referents.On the other hand, the demonstrative pronoun esto 'this' shows a strong preference to be combined with present tenses in our corpus sample.It is surprising the low frequency of [esto + past tense] combinations, especially if we consider the overall high frequency of this demonstrative.Demonstrative eso 'that' also shows a strong preference to be combined with present tenses in the corpus (0.81 %) but it doubles esto figures as regards past tense combinations (0.19% and 0.09%, respectively.)Thus, a tendency can be observed in the poles of the tripartite demonstrative system analyzed according to which proximate esto shows a preference to co appear with present tenses while distal aquello shows a preference to co-appear with past tenses in discourse.On the other hand, medial eso shows a more balanced behavior and, while showing a preference for present tenses in our corpus sample, it is less constrained to co-appear with past tenses than proximate esto.As I mentioned above, our demonstrative pronouns can be used in two different ways principally.They can be used in a deictic proper use to refer to an entity in the extra-linguistic context like or in a discourse deictic use to refer to some textual entity, that is, something that has been previously mentioned in the discourse.In deixis proper, spatial distance is important insofar the speaker relies on the spatial dimension to pick up the most adequate demonstrative, hence the relevant tripartite distinction proximate-medialdistal.In discourse deictic uses, on the other hand, it is commonly assumed that demonstratives are simply voided of such spatial dimension or that the spatial dimension is reanalyzed into alternative textual or discourse dimensions.In my opinion, this may be happening with demonstrative pronouns in modern Spanish.The data presented in this paper suggest that Spanish speakers may be reanalyzing the spatial deictic dimension (distal ~ proximate) into a tense, or time-related, dimension (present ~ past) when intra-textual (i.e.endophoric) reference is involved.We have seen that the distal demonstrative pronoun aquello is strongly constrained by past reference contexts, which is supported by our corpus frequencies of [aquello + tense].Conversely, our corpus sample indicates that the proximate demonstrative pronoun esto shows a strong tendency to be combined with present tenses in discourse.Medial eso would be unmarked with respect to distance or temporal dimension, which is supported by the fact that medial pronoun eso is the most frequently used demonstrative pronoun in Spanish across all varieties (see Tables 3 and 5).How speakers reanalyze or recategorize spatial deixis systems into temporal deixis systems is beyond the purposes of this paper.
A third corpus study was carried out to check the behavior of demonstrative determiners and compare it with the demonstrative pronouns.To that purpose I analyzed 150 occurrences of the combinations este hombre, ese hombre and aquel hombre 'this/that/that man': the three masculine singular demonstrative determiners with a noun phrase that denotes a physical entity.The main goal was to check if the same constraints that we have seen for pronouns could also be applied to demonstrative determiners.The reason for having chosen a noun like hombre 'man' was to check whether the particular denotation of the referring expression involved might also have an effect in the interaction tense-demonstration.The demonstrative noun phrases were analyzed to see whether they were accompanied with a present or a past tense.Two examples of each demonstrative (1 case with present tense, 1 case with past tense) are shown in ( 15)-( 20) to illustrate the third corpus analysis.
(15) "El agresor nos decía que aquel hombre había atacado a varios amigos suyos durante meses", explicaba un marine a una cadena de televisión.'The aggressor told us that that man had attacked some of his friends for months', a marine explained to a television channel.' (La Vanguardia, 1994) ( 16) De repente, la voz aterrorizada de uno de los pescadores sobresalta a todos.Espadarte!.Todos se vuelven hacia donde señala aquel hombre.'Suddenly, the terrified voice of one of the fishermen alarmed everyone.Espadarte!Everybody turned towards where that man was pointing.'(Duelo en el Mar, 1995) (17) El protagonista es un locutor, Pablo, que está enamorado de una chica que fue su amor de juventud, pero está casado.Yo soy la hermana del marido.De alguna manera es la relación de ese hombre con tres mujeres muy distintas.'The main character is an anchorman, Pablo, who is in love with a girl who once was his puppy love, but he is married.I am the husband's sister.In some way, it is about the relationship of that man with three different women.' (El Diario Vasco, 2004) (18) Fue tan injusto y brutal el fusilamiento de ese hombre bueno que fue Carrasco, que todavía sería mayor la injusticia si para ensalzar su figura tuviera que … 'The execution of that good man named Carrasco was so unfair and brutal, that such unfairness would even be bigger if, to glorify his figure, he had to…' (La Vanguardia, 1995) (19) El problema de este señor es que sufre una especie de enfermedad mental contra el alcalde de Barbate.Además, este hombre ni es promotor porque termina sus casas con muchos problemas e incumpliendo los plazos de sus inversiones.'The problem with this guy is that he suffers some kind of mental illness against the major of Barbate.Furthermore, this man is not even a developer because he builds houses with many issues and fails to meet the deadlines of his investments.' (Tiempo, 1990) (20) Entre la concurrencia circuló el rumor de que este hombre había ido a quejarse de que el reloj de coleccionista no le funcionaba.'Among the audience, the rumor spread that this man had complained that his special collector's watch didn't work.' (La Vanguardia, 1994) The results of the study on determiners are shown in Table 12.Our two variables are demonstrative determiners and tense.The category Other includes those tokens which could not be included in either Present or Past because they were accompanied by, for example, a conditional or a future tense.The figures in this table are similar to the distribution for demonstrative pronouns observed in Table 8.We can see the strong predilection of distal demonstrative aquel for past tenses and really low numbers for the same demonstrative in combination with present tense in the analyzed examples.Proximate este and medial ese show very similar results, with an overall strong preference for present tenses over past tenses although proximate este though shows even a stronger preference for present tenses than medial ese (0.64% versus 0.58%.)It is also interesting that medial ese is more frequently found combined with other tenses than este or aquel (n = 9 in category Other).This might be another indication that ese (like its counterpart eso) may be acquiring a more neutral or unmarked character regarding reference in discourse.The relative frequencies of demonstrative determiners and tense combinations are shown in Table 13 and graphically in Figure 2.

Contextual licensing
Our corpus study has revealed an abundance of cases where demonstrative aquel is used in discourse along with a past tense accompanying the demonstrative or in its immediate textual surroundings.The number of cases is too high to be due to sheer chance, so there must be some linguistic mechanism that is responsible for such connection.I will argue that the linguistic element responsible for the connection tense-demonstration is a linguistic expression that denotes a past time, which is to be found in the textual environment of the distal demonstrative aquel.Thus, the presence of this element triggers or licenses the pragmatically felicitous use of the distal demonstrative as a discourse anaphor in both adnominal and pronominal form.Grammatical tense expresses the time at, during, or over which the state or action denoted by the verb occurs.In this respect, past tenses are perfectly suited to act as licensors of the distal demonstrative aquel in discourse and, in fact, past tenses are the main contributors to the congruence tense-demonstration in Spanish.But there are cases in which the congruence is not so transparent since the alleged correlation tensedemonstration does not seem to be solely restricted to the category of grammatical tense.Cases abound where the distal demonstrative aquel is not accompanied by a past tense within the same clause nor even in its most immediate vicinity.A suitable explanation must be provided for these and similar cases since, otherwise, the validity of our study could not be demonstrated and the hypothesized congruence tensedemonstration in Spanish could not be sustained.Consider example ( 21).
(21) El 14 de abril de 1931 se proclama la II República Española.El día 15 Franco da conocimiento a la Academia General Militar de aquel acontecimiento patrio, con estas palabras: "Proclamada la República en España, …" 'The II Spanish Republic is proclaimed on April 14 1931.On the 15th, Franco reports that national event to the General Military Academy with these words: "Having proclaimed the Republic in Spain …" (Prensa española, 1977) Not a single past tense can be found in ( 21) but, instead, we find several present tenses used with a past meaning.These are cases of the so-called historical present: the use of present tenses to narrate past events.Historical Presents challenge the hypothesis of the congruence tense-demonstration because no past tenses show up, however the distal demonstrative aquel is still licensed in the discourse.Examples like (21) and similar ones clearly indicate that the mechanism responsible for licensing the demonstrative does not entirely rely on grammatical tense so I will rather advocate for conceiving the congruence phenomenon as semantic or pragmatic in nature.The hypothesis that I will present in this paper regarding the congruence tense-demonstration is that contextual clues like tense but also temporal adverbials, noun phrases and other temporal expressions favor the use of certain demonstrative anaphors.More specifically, I will argue that distal demonstrative aquel ('that yonder'), both in its pronominal and adnominal forms, requires a PAST TIME contextual element to be felicitously licensed in discourse.The crucial point is that not only tense as a grammatical category or the events denoted by the verbal material are responsible for the particular behavior of distal aquel but rather the more abstract category of TIME and more specifically PAST TIME the underlying factor explaining the observed co-occurrence.Conversely, the two other demonstratives of the tripartite system (proximal este ('this') and medial ese ('that') do not need any particular contextual configuration to be used as demonstrative anaphors in discourse or, in other words, they do not require any particular time-expressing contextual trigger to be fully licensed and they can be freely used by speakers regardless of the temporal frame in which the discourse is set up.Thus, in (21) the element licensing the use of distal aquel is the temporal expression el 14 de abril de 1931, a calendar date that serves to locate the narrated event (and most likely the discourse topic) at some past time prior to the speech time.This example clearly illustrates the convenience to appeal to the abstract notion of time to explain the congruence phenomenon.By PAST TIME here, I mean the ability of certain linguistic elements (tense, adverbials, etc.) to situate events, states, actions or situations at a particular point in time (present, past or future) relative to some other time (most commonly the speech time) or to even directly denote a specific point in time (calendar dates, etc.). 11The category PAST TENSE is restricted to verbs (i.e. the way time, present, past or future, is marked via morphology in the verb) so it is not enough to explain the congruence in cases like (26).Again, the expression el 14 de abril de 1931 is the key that establishes a past time frame of reference relative to the speech time.The speaker, by using that expression, makes clear that the following sequence of narrated events are to be interpreted relative to the time conveyed by the expression itself, that is, at a past time prior to the speech time t ≺ n.
Unfortunately, not all cases of historical present can be given an explanation along the lines of (21).Consider example ( 22). ( 22) Durante toda su vida, Napoleón Bonaparte vive soñando con conquistar Europa en poco tiempo.Sin embargo, en aquella época, la tecnología militar no permite avances rápidos.'During all of his life, Napoleon Bonaparte dreams with conquering Europe in a short time.Nevertheless, in that time, the military technology does not allow for fast advances.' Example ( 22) is a case of historical present with no temporal expression that may contribute to establish a temporal frame of reference for the narrative; but still the use of the distal demonstrative noun phrase aquella época is fully felicitous.It may be argued that the demonstrative expression itself is enough to interpret the historical present tenses as past tenses.In this line of reasoning, a past tense interpretation for the present tense would be coerced upon processing of the distal demonstrative.This hypothesis is appealing for it would allow us to explain cases of historical present by just appealing to the grammatical category of [PAST TENSE] and discarding the more abstract notion of [PAST TIME].Accepting the coercion hypothesis would also allow us to explain the historical present in (21) but somewhat differently.In that case, the temporal expression el 14 de abril de 1931 would be the element that would coerce a past interpretation for the historical present tenses and, once given a past interpretation, the present tense(s) would become an appropriate [PAST TENSE] licensor for the distal demonstrative noun phrase aquel acontecimiento.The main problem with the coercion hypothesis is that it does not allow us to explain cases where no historical presents or even past tenses whatsoever are involved (i.e.examples ( 23) and ( 24) below) unless, of course, we are willing to accept that future tenses also get a past interpretation via coercion.In any case, when no other temporal clues are present, historical presents are problematic for the congruence phenomenon as no overt or lexical time or tense feature can be said to act as licensor of the distal 11 Regarding the theoretical status of TIME, I believe it necessary to explain the ideas concerning the congruence tense-demonstration defended in this paper.By TIME here I mean a specific temporal point (PAST, PRESENT or FUTURE) at which some event or situation is anchored and which is expressed by some linguistic expression.Different expressions can convey TIME differently.Thus, for example, a date denotes a specific point in time (i.e.April 25th 2000) and can thus be said to express TIME directly.Tenses, on the other hand, express TIME indirectly by situating an event at a certain point on a time plane relative to the speech time.The convenience of proposing the category TIME and, more specifically, PAST TIME is that it allows us to subsume all these expressions under one single category.
demonstrative.In such cases, we can only appeal to accommodation as an explanatory mechanism or to insufficient context to find the appropriate licensor. 12 In order to test whether the semantic nature of the entity referred to or the specific demonstrative expression (pronominal vs. adnominal) may have an effect on the clear connection tense-demonstration shown by demonstrative aquel, I analyzed a few occurrences of the distal demonstrative determiner aquel and pronoun aquello from our corpus.The main goal was to test whether other explicit contextual factors besides tense may have an influence in the licensing of this element.To that aim, I scrutinized a substantial piece of text surrounding the demonstrative anaphor (a piece of text consisting of an average of ten sentences).Our hypothesis that the time at which the discourse is setup greatly contributes to the use of demonstrative aquel seems to be confirmed upon a close analysis of these examples.In ( 23), there are no past tenses but only future tenses instead.I will argue that the noun phrase las misiones Apolo ('the Apollo missions') is the element that sets up the narration time hence licensing the use of the distal demonstrative.While it is true that there is no overt temporal expression, NPs denoting past events are very commonly found accompanying the distal demonstrative in discourse.In these particular cases, world or background knowledge is crucial for the participants to interpret the discourse as referring to a past event.But cases like (23) involving noun phrases that can only be interpreted as past events by appealing to the world knowledge of the interlocutors are problematic for a formal characterization of the congruence phenomenon that will be presented in § 5 in terms of Temporal Discourse Models.The problem arises from the lack of TIME as a lexical feature that may be responsible for licensing the distal demonstrative.Based on world knowledge it is clear that a noun phrase like Las Misiones Apollo denote a series of events that occurred in the past but, to the best of my knowledge, there is no way to encode the past reference of these NPs in our TDMs.
(23) Este será el legado histórico de las misiones Apolo.Por esto será por lo que aquel hecho singular entrará por méritos propios en la historia de las hazañas más importantes de nuestra civilización.'This will be the historical legacy of the Apollo missions.This will be the cause why that singular fact will enter on its own merit the history of the most important accomplishments of our civilization.'(Hombres en el espacio.Pasado, presente y futuro, 1996) A range of temporal expressions can license the use of distal aquel.In (24), no past tenses occur with the demonstrative expression aquel hecho ('that fact') in the last sentence.The demonstrative refers back to the event that is described in the immediately previous sentence Que el año pasado … in a clear discourse anaphoric use.In fact, only a past tense in the technical sense (the Spanish pretérito perfecto compuesto 'ha amedrentado') can be found in this discourse and this is a past tense, 12 I would like to thank an anonymous reviewer for pointing out to me is that historical present contexts can always be substituted by a preterite but never by an imperfect, which may bring additional evidence to the distinction between the preterite as deictic (i.e. it must be interpreted relative to some contextual parameter) and the imperfect tense as anaphoric (i.e. it can only be fully interpreted in relation with other event).Thus, a simple contextual shift is enough to fully interpret the preterite, but it is not enough to interpret the imperfect in historical present contexts.A comparison between these two tenses may certainly contribute to unveil the mechanisms underlying the contextual licensing that we propose in this paper (i.e.contextual licensing may not be (only) presuppositional but it may be comparable to the way indirect speech shifts the landmarks of the utterance.)This idea will be worth exploring in future extensions of this paper.
which is commonly used to convey the current relevance of the denoted event/situation.Two imperfect tenses in the subjunctive mood and an instance of the Spanish copula estar ('to be') in the present are all the tenses used in this example.There is, though, a temporal expression that clearly contributes to set the narrative at some time in the past, namely, el año pasado ('last year').
Que el año pasado un alumno de 16 años disparara en mitad de un examen de matemáticas una escopeta de caza y el tiro pasara a metro y medio del profesor no ha amedrentado a nadie.De aquel hecho, ahora solo queda una gran satisfacción.
'The teachers of Móstoles IV high school are living in a hustle and bustle these days.The fact that a 16 year old student shot a hunting rifle in the middle of a math exam last year and the shot passed over the teacher at one and a half meters distance didn't scared anybody.Only a great satisfaction remains from that fact.' (El País, 01/10/1988) The discourse in ( 25) combines the use of past tenses along with demonstrative anaphoric reference to the noun phrase the French Revolution.A past tense also cooccurs with the distal demonstrative in the last sentence.It appears that the NP the French Revolution is the element that licenses the use of the distal demonstrative.Thus, I'll argue that, in cases like (25) and similar ones, world knowledge is crucial for the hearer to interpret the use of the distal pronoun when other contextual clues are absent.In this particular case, the hearer must be able to interpret the French Revolution as a noun phrase that refers to a past event.
(25) El editorial que publicó El País el pasado día 19 de Julio lo podría firmar cualquier diario conservador.Sólo le ha faltado una arenga anticomunista y una referencia a lo utópico de las revoluciones.En Nicaragua, y usted lo sabe bien, se ha avanzado mucho en lo que se refiere a libertades y a igualdades.El bicentenario de la Revolución Francesa parece que no ha servido ni siquiera para dejar claro cuál es la más importante de las palabras que encabezaron aquel hecho histórico.'The leading article, published by El País last July 19th, could be signed by any conservative newspaper.It was only in need of an anti-communist harangue and a reference to the revolutionary utopia.In Nicaragua, and you know that well, much progress has been made regarding liberty and equality.It appears that the bicentennial of the French Revolution was not even useful to make it clear which word is the most important one among those leading to that historical fact.' (El País, 01/08/1989) In summary, at least one past-triggering element could be found in all the examples analyzed thus far.In many cases, more than one past trigger co-appeared in the anaphor's immediate context.What all these examples have in common is that they all feature some linguistic expression that either directly or indirectly denotes a past time hence contributing to positioning the event talked about at some point in the past of the speech time.Cases of the distal demonstrative aquel in Peninsular Spanish which do not feature at least one of the mentioned PAST TIME linguistic expressions are very hard to find in written and spoken discourse so, in our opinion, this might be a clear indication that distal demonstrative aquel is strongly favored in contexts where a past temporal frame has been established via past tense, temporal expressions (dates, adverbials, etc.) or lexical material (event denoting NPs and perhaps other lexical elements).Conversely, cases of the demonstratives este and ese in discourses featuring no past tenses whatsoever or any other contextual clues that may set the reference time at the past of the speech time are not infrequent.The next three examples illustrate this point.In ( 26), the demonstrative noun phrase este hecho ('this fact') anaphorically refers to factual information described in the previous clause.The event or fact talked about set forth in the first clause is clearly hypothetical as it is described as something that is going to happen today but not yet occurred.Surprisingly, the only temporal expression in ( 26) is the adverb hoy ('today') that clearly helps in establishing the reference time at a point that includes or overlaps the speech time.Note also that no past tenses appear in the text.The use of the expression aquel hecho ('that fact') would have been clearly infelicitous in this context.
'No debates or voting are planned for today's session.For some of the party leaders, this fact makes the conference "a mere act of cosmetics" and it prevents them to discuss in depth the party's programmatic and strategic target.'(La Vanguardia, 12/10/1994) Finally, the next two examples show that discourses in which reference is made to a past time or past event are not exclusive of the distal demonstrative aquel.In (27), we find the time expression en agosto de 1992 and a series of past tenses that combined contribute to set the reference time of the event described at some point in the past of the speech time.In these two examples, the speaker/writer uses the medial and proximal demonstratives ese and este, respectively.Again, these examples help to confirm that any context regardless of its temporal frame licenses the use of these demonstratives.
(27) Los terroristas pretendían asestar ayer un golpe para añadir a la cadena de atracos que iniciaron en agosto de 1992, cuando intentaron robar infructuosamente un furgón en León.A ese hecho siguieron otro asalto a otro furgón en Santander.'Yesterday, the terrorists tried to strike one more blow to add to the chain of assaults that they began in August 1992, when they unsuccessfully tried to steal a van in León.That fact was followed by one more assault to another van in Santander.'(La Vanguardia, 15/11/1994) As we have seen in this section, linguistic expressions that encode a past time like calendar dates, adverbs ('tomorrow' 'today') and adverbial phrases ('last year') and a range of lexical material contribute to situate the event under discussion at a particular point in the past of the speech time.In some cases the time reference can be as definite as with a calendar date (i.e.April 13th 1890) or indefinite as with past tenses.In other cases, the world knowledge of the discourse participants plays an important role in determining the temporal reference as with lexical items like the Beijing Olympic Games or the assassination of President Kennedy.In cases like these, it is the world knowledge of the interlocutors or their common ground what contributes to anchor the referent of the noun at a particular point in the past of the speech time (i.e. the summer of year 2008 and year 1963, respectively.)Thus, if we assume that common ground and a certain amount of shared world knowledge is essential to communication, some specific noun phrases may also contribute to the setup of the referential temporal frame of discourse.Finally, tense is another key contextual factor that contributes to the setup and maintenance of the temporal frame of the discourse.In many cases, the past tense verb form(s) can be found at a distant position in the text from the location of the demonstrative anaphor itself.In other cases, the past tense is found in close syntactic connection with the anaphor (i.e. the demonstrative is the subject of the verb).A good number of the cases analyzed in this section are narratives, where some past fact, event or situation was described by making use of diverse past tenses along a variable textual span.We have seen that the Spanish distal demonstrative aquel in discourse anaphoric uses, and perhaps other demonstrative anaphors in different languages, appear to be highly sensitive to the temporal information conveyed by certain expressions up to the point that the demonstrative needs a past denoting 'trigger' in its textual surroundings to be fully licensed in context.The results of the corpus study in § 3 show an interesting preference for the proximate demonstrative este/esto to appear with present tenses.The medial demonstrative ese/eso showed no particular preference for a specific tense so it seems that this demonstrative is used regardless of the temporal frame in which the discourse is set up.

Temporal discourse models and the connection time-demonstratives
Most of the examples analyzed in this paper represent uses of demonstrative anaphors in narrative discourse.This is not surprising as making reference to past events commonly implies that the event in question is being described as part of a larger sequence of events, that is, a narrative.I think that the structure of narratives may help us characterize the congruence that is subject to analysis in this paper.In order to provide a formal account of the correspondence time-demonstration shown by demonstrative aquel I will make use of the Temporal Discourse Models for Narrative Structure (TDMs, henceforth) developed by Mani & Pustejovsky (2004).TDMs constitute an adequate way of characterizing the order of happening of events in the narrative while serving as representations of the temporal structure of narratives.Such temporal models will allow us to represent the mechanism underlying the special relationship that commonly arises between the distal demonstrative aquel and past time expressions in Iberian Spanish.
A TDM is a tree-structured syntactic model of global discourse structure, where temporal relations are used as substitutes for discourse relations, and where abstract events corresponding to entire discourses are introduced as nodes in the tree.The narrative models are tree-structured where the dominance relation is expressed by temporal inclusion (i.e. a node X is a child of node Y if and only if X is temporally included in Y).Consider the narrative discourse (28) and the corresponding tree T1.13The discourse or narrative in (28) has been divided into its constituent sentences (a) through (e), each of these describing an event.Regarding the tree representation (T1) for ( 28), E0 has children Ea and Ee, Ea has children Eb and Ec, and Eb has children Ed.The nodes with alphabetic subscripts are events mentioned in the text (Ea, Eb, Ec, etc.), whereas nodes with numeric subscripts are abstract events (E0), i.e., events that represent abstract discourse objects.Thus, E0 is an abstract node representing a top-level story.But not only abstract events create embeddings.Textually mentioned events can also create embeddings.In (28), the event Ea creates an embedding that includes events Eb and Ec and the event Ec, in turn, creates an embedding that includes Ed.
TDM representations like T1 are semantic representations of event nodes so not necessarily isomorphic with the structure of the narrative at the text level.Also, although events are ordered left to right no temporal ordering is directly represented in the tree.What is directly represented in the tree is event inclusion.Thus, for example, we can say that event Ea (Edmond made his own Christmas presents this year) includes subevents Eb, Ec and Ed in a way similar to the topic-subtopic relation or the rhetorical relations narration or elaboration in the sense given to these relations in Asher & Lascarides (2003).
TDMs also feature a set of temporal ordering constraints.The set of temporal ordering constraints for T1 is the following: C1 = {Ea ≺ Ee, Eb ≺ Ec, Eb ≺ Ed, Ec ≺ Ee, Ed ≺ Ee}.As the authors point out (2004: 2): "A TDM is a pairing of tree structures and temporal constraints.More precisely, a Temporal Discourse Model for a text is a pair <T, C > where T is a rooted, unordered directed tree with nodes N = {E ∪ A}, where E is the set of events mentioned in the text and A is a set of abstract events, and a parent-child ordering relation ⊆ (temporal inclusion).A non-leaf node can be textually mentioned or abstract.Nodes also have a set of atomic-valued features.Note that the tree is temporally unordered left to right.C is a set of temporal ordering constraints using the ordering relation '≺' (temporal precedence) as well as minimal restrictions on the above temporal inclusion relation (expressed as ⊆ min).
'The teachers of Móstoles IV high school are living in a whirl these days.That a 16 year old student shot a hunting rifle in the middle of a math exam last year and the shot passed over the teacher at one and a half meters distance didn't scared anybody.Only a great satisfaction remains from that fact.'The properties of each individual node in T2 are the following:14 Node.properties(Ea): live_in_whirl(Ea, x), the teachers(x), high-school(y) these_days(t 1 ), Loc(Ea): t 1 , Ea ⊆ n Node.properties(Eb): fire(Eb, s, g), student(s), rifle(g), last year(t 2 ), Loc(Eb):t 2 , Eb ≺ n Node.properties(Ec): miss(Ec, b, j), the bullet(b), the teacher(j), t 3 = t 2 , Loc(Ec): t 3 , Ec ≺ n Node.properties(Ed): remain_from(Ed, fact(Eb∧Ec), r), great relief(r), now(t 4 ), Loc(Ed): t 4 , fact(Eb∧Ec), Ed ⊆ n TDMs can be given different representations such as temporal first order logic representations or as a Discourse Representation Structures (DRS, henceforth) by expanding the properties of the nodes with predications about the narrative events under consideration.A DRS representing the narrative discourse ( 29) is shown in Figure 5.Our DRS allows us to establish a set of discourse referents (the universe of the discourse) in the upper part of the box {Ea, x, y, Eb, s, g, etc.} and a set of conditions on those referents {live_in_whirl(Ea, x), the teachers(x), etc.}The set of discourse referents and the conditions on them has been set to a minimum to avoid complexity and facilitate the reading.TDMs represent events as pairs of time points (time intervals) and events are treated as primitives in Discourse Representation Theory (Kamp and Reyle 1993). 15Thus, time intervals are introduced in a DRS in the form of e ⊆ t (for events) and e Ο t (for states).Reference times are represented with variables t 1 , t 2 … t k .
In the DRS in Figure 5, the referent of the demonstrative ese hecho is the event represented as fire (Eb, s, g).As specified in the DRS, the event referred to (Eb) is temporally located at t 2 (i.e.Loc(Eb): t 2 ), included in reference time t 2 (Eb ⊆ t 2 ), and t 2 is a time prior to the utterance time n (t 2 ≺ n).Succinctly, t 2 is a past time and the event described and referred to with the demonstrative is a past event included in t 2 .Note that grammatical tense as such is not overtly represented in our DRS, hence the condition that serves to establish a past reference point for the event Eb is the expression last year in the second sentence of ( 29) and introduced as a DRS condition 15 In Kamp & Reyle (1993) where: a. U K ⊆ V (the set of discourse referents in K is included in the set of variables V). b.Con K is the set of conditions in K.The members of Con K can be atomic conditions or complex conditions.
i.An atomic condition is a condition of the following kind: 1. x = y, where x, y ∈ V. 2. P(x), where x ∈ V and P is either a proper name, a common noun or intransitive verb.3. P(x, y) where x, y ∈ V and P is a transitive verb.ii.A complex condition is a tuple of the following kind: where K 1 and K 2 are DRSs.Besides DRSs construction conditions there are also embedding conditions and verifying conditions of DRSs in a model.The embedding conditions are formulated formally as: Be K a DRS and S a set.An possible embedding of K in S is a function f such that: a.The range of f is a subset of S (Rg(f) ⊆ S). b.U K is a subset of the domain of f (U K ⊆ Dom(f)).c.If a discourse referent x appears in K but not in U K , x is not an element of the domain of f.A function f' is an extension of f if (i), (ii), or both, obtain: (i) Dom(f) ⊆ Dom(f'); (ii) Rg(f) ⊆ Rg(f').Finally, be K a DRS, M a first order model, f a possible embedding of K in U M (the Universe of the Model).Function f is a proper or verifying embedding of K in M iff: a. f verifies every atomic condition γ then every extension of f which is a proper embedding of K 1 in M can be extended to a function which is a proper embedding of K 2 in M. d.If [K 1 ∨ K 2 ] ∈ Con K , then either there is an extension of f which is a proper embedding of K 1 in M or there is an extension of f which is a possible embedding of K 2 in U M and a proper embedding of K 2 in M.

last_year(t 2 ).
The reference of the demonstrative noun phrase ese hecho is the conjunction or composite of the two preceding events Eb and Ec, expressed in the DRS with the condition that_fact(Eb∧Ec).In summary, the DRS representation of the TDM allows us to formally characterize the temporal structure of the narrative, to resolve the referential link between the demonstrative anaphor and the event referred to as well as precisely locate the time expression that sets the narrative at a particular reference point in the past and which licenses the use of the distal demonstrative aquel.Ea, Eb, Ec, Ed, x, y, n, s, g, b, j, r, t 6) is decomposed into its constituent propositions (a) through (d) in (30).The corresponding tree (T3) is shown in Figure 6 and the set of ordering constraints for T3 is the following: C3 = {Eb ≺Ea, Ec≺Ea, Ed≺Ea, Eb⊓Ec, Eb ≺Ec, Ec≺ Ed}.
'The terrorists tried to do a job yesterday.'b.Los terroristas iniciaron una cadena de atracos en agosto de 1992.
'The terrorists started a chain of assaults in August 1992.'c.Los terroristas intentaron robar infructuosamente un furgón en León en agosto de 1992.'The terrorists unsuccessfully tried to steal a van in León in August 1992.'d.A ese hecho siguió otro asalto a otro furgón en Santander.
'That fact was followed by one more assault to another van in Santander.'The properties of each individual node in T3 are the following: Node.properties(Ea): try_to_do(Ea, x, y), terrorists(x), job(y), yesterday(t 1 ), loc(Ea):t 1 , Ea ≺ n Node.properties(Eb): start(Eb, x 1 , j), chain_of_assaults(j), x 1 =x, August 1992(t 2 ), loc(Eb):t 2 , Eb ≺ n Node.properties(Ec): try_to_steal (Ec, x 2 , v), van(v), x 2 =x, t 3 =t 2 , loc(Ec):t 3 , Ec ≺ n Node.properties(Ed): assault(Ed, x 3 , k), another_van(k), x 3 =x, t 4 , loc(Ed):t 4 , t 4 ≻ t 3 , fact(Ec), follow(Ed, Ec), Ed ≺ n, Ed ≻ Ec Like discourse (29), discourse ( 30) is composed of four different propositions, each of them conveying a different event Ea through Ed.The discourse as a whole is framed in the past and all the individual events temporally located in the past of the utterance time.This is indicated by the temporal location conditions in the node properties and in the DRS for discourse (30) shown in Figure 7 (i.e.Loc(Ea): t 1 , t 1 ≺ n, Ea ⊆ t 1 , Ea ≺ n), and the temporal location information is linguistically expressed by means of different temporal expressions (i.e.yesterday, August 1992) and by inferring temporal relations among events from the meaning of certain verbs (i.e.follow (Ed, Ec).Note also the consistent use of past tenses throughout the text.All these linguistic elements give us the necessary information to interpret the discourse as a whole and its component events as framed at some time in the past of the utterance time.These are the type of contexts that most commonly license the use of the temporally marked distal demonstrative 'aquel' as we have described and illustrated in previous sections.The medial demonstrative noun phrase ese hecho ('that fact') appears in the last sentence and refers back to the event described in Ec.As we know, the medial demonstrative 'ese' is completely unmarked regarding time, so it is also licensed in past-framed contexts like (30).The speaker refers to the event described in sentence (c) as a fact instead of an event, so in the mind of the speaker the event described shifts or becomes factual information once it is processed.The condition fact(Ec) in the DRS expresses this idea, and the condition follow(Ed, Ec), along with the temporal conditions, indicate that the assault event Ed followed the event, or fact, Ec.Knowing that the temporal location of the event Ed is t 4 and the logical relation among events overtly expressed with the verb 'follow' we can therefore infer that Ed succeeds Ec or Ed ≻Ec.Ea,Eb,Ec,Ed,x,y,x 1 ,x 2 ,x 3 ,j,v,n,k,t 21) is decomposed into its constituent propositions (a)-(b) in (31).The simple tree (T4) for this short discourse and the corresponding DRS are shown in Figures 8 and 9, respectively.The set of ordering constraints is C4 = {Ea ≺ Eb}.This discourse is particularly interesting for it is narrated using the historical present but, at the same time, it is a very specific temporal expression, a calendar date, what gives us the only overtly expressed clue to understand that the narration is framed and has to be interpreted as located in the past.
(31) a.El 14 de abril de 1931 se proclama la II República Española.'The II Spanish Republic is proclaimed on April 14 1931.' b.El día 15 Franco da conocimiento a la Academia General Militar de aquel acontecimiento patrio, con estas palabras: "Proclamada la República en España..." 'On the 15th, Franco reports that national event to the General Military Academy with these words: "Having proclaimed the Republic in Spain …' The properties of each individual node in T4 are the following: Node.properties(Ea): be proclaimed(Ea, x), Spain's II Republic(x), April 14th 1931(t 1 ), Loc(Ea):t 1 , Ea ≺ n Node.properties(Eb): report (Eb, y, Ea, m), Franco(y), Military Academy(m), April 15th (t 2 ), Loc(Eb):t 2 , Eb ≺ n Despite the absolute lack of past tenses in this DRS, the temporal clue that allows us to interpret the two events as preceding the utterance time is the calendar date that is introduced in the first sentence (i.e. April 14th 1931).This date established the discourse time in the past, hence the conditions Ea ≺ n and Eb ≺ n.The world knowledge of the interlocutors might have played a role in this narration due to the introduction in the discourse of the proper names Franco, the Spanish dictator, and the II Republic, but the overt temporal expression in the first sentence makes unnecessary to employ world knowledge to temporally locate the component events.This temporal context clearly licenses the use of the temporally marked distal demonstrative aquel in the second sentence.The NP aquel acontecimiento patrio ('that patriotic event') clearly refers back to the first component event Ea, which is a past event, and the referential link is shown in the DRS with the condition that_patriotic_event(Ea).Ea, Eb, x, y, m, n, t

Demonstratives as anaphoric presupposition triggers
The remarkably low frequency of occurrence of the distal demonstrative aquel/aquello shown by our corpus study and its rather constrained use (i.e. it is licensed in contexts that require a past reference frame setup when reference is made within the text) make the aquel/aquello a marked term in the tripartite system of Spanish demonstratives.On the other hand, our corpus study indicates that the proximate demonstrative este/esto shows a preference to be combined with present tenses in discourse and that the medial demonstrative pronoun ese/eso does not appear to show any temporal constraints.
Additional evidence that supports the status of demonstrative aquel/aquello as highly marked term in the tripartite system comes from the neutralization phenomena between the forms aquel/aquello and the definite article el (and the neuter lo), which may be a sign that this demonstrative is developing new meanings that move it away from the two other elements.Thus, for example, distal aquel shows some parallel uses with the definite article that do not share with the proximate demonstratives. 16 Based on the temporal constraints that impose restrictions in the use of demonstratives in Spanish discourse, a summary of their features in the connection tense-demonstration is provided in Table 14.The distal demonstrative aquel/aquello 16 Historical considerations concerning the genesis and evolution of the Spanish demonstratives from Latin may be considered additional support in favor the weakening of distal aquel as a demonstrative.The modern definite article in the Romance languages evolved from the Latin distal demonstrative ILLE (Span.Aquel≺ Lat.ILLE) and not from the other demonstratives HIC ISTE or IPSE.Perhaps, the modern distal demonstrative aquel of Iberian Spanish is going through a similar evolutionary path than its Latin counterpart.
is the marked element of the tripartite system due to the strong constraints related with time (past).On the other hand, medial ese/eso is an unmarked element for it shows no particular constraints related with tense or time.Finally, the proximate demonstrative este/esto shows an intermediate status regarding markedness.As the data from the corpus indicate, este/esto appears to be inclined to co-occur with present tenses rather than with past tenses.This might be an indication that the proximal demonstrative is time constrained (with present) or that present time frames of reference license the use of this demonstrative in a higher proportion than past time frames of reference in discourse.Our data may not be conclusive because the proximal demonstrative can also be combined with past tenses in a relatively high proportion, hence my considering este/esto undefined (+/−) as regards markedness and time constrained.The summary is as follows.Given that demonstratives per se have little descriptive content, several authors working on the semantics and pragmatics of demonstratives have characterized these elements in terms of the presuppositions that they trigger.Zeevat (1999) identified the following basic pre-conditions or presuppositions triggered by the use of a demonstrative: a presupposed utterance, an agent (i.e. the utterer of the demonstrative expression), a time (i.e. the utterance time), a proximity condition to distinguish among distal and proximate demonstratives and, arguably, a presupposed pointing (i.e.not all demonstratives are accompanied by an overt pointing or ostension).This author developed presupposition discourse representation structures (presupposition DRSs) to characterize said content.
As an extension of her previous work on pronouns, Roberts (1993) gives demonstratives a treatment akin to the one given for pronominal elements and definite descriptions.For her, what makes demonstratives special -either in a determiner or pronominal use-will be a particular presuppositional feature that unlike pronouns and definite descriptions demonstrative expressions have, i.e. a demonstration presupposition.Furthermore, she will also assume that this presupposition of a demonstration is general for all uses of demonstratives and not only to those uses involving a deictic gesture.Roberts starts by assuming that all demonstratives are definite, that is, they carry presuppositions of familiarity and informational uniqueness and they, furthermore, carry a demonstration presupposition. 17t is important to note that Roberts' account of definite NPs -let's recall that demonstratives are all definite in her theory-constitutes the presuppositional counterpart of Russell's (1905) logical form for definite description.Thus, all definite NPs will have an existence presupposition and a uniqueness presupposition which are to be taken as the counterparts of the Russellian existence and uniqueness.Contra Russell's, the existence and uniqueness of definites are part of the presuppositional content of these expressions and not of their asserted content Crucial for Roberts' theory of definite expressions and, consequently, for her theory of demonstrative NPs is the assumption that the existence of definites is informational uniqueness.What this means is that we won't be considering existence as existence of an individual in a model but existence of a discourse referent in the domain of discourse "Use of a definite description does not entail that there is some entity in the model/world which uniquely bears the descriptive content of the NP, but only that there is a familiar discourse referent in the context which is the only element in the Domain that is entailed by the common ground to bear the NP's descriptive content".Thus, uniqueness has to be understood as informational uniqueness and not as uniqueness in the world.
Another critical concept in Roberts (2003) work on demonstratives is that of familiarity.There are two basic types of familiarity for a discourse referent that license its introduction into the interlocutors' common ground: strong and weak familiarity.The latter can be, in turn, subdivided into three subtypes.

•
Strong Familiarity: the discourse referent which serves as the antecedent of the NP has been introduced in the discourse representation via utterance of a preceding NP.

•
Weak Familiarity: (i) The discourse referent is familiar to the interlocutors either through perceptual acquaintance or simply for belonging to the common general culture.(ii) The discourse referent is contextually entailed.(iii) The discourse referent has to be accommodated or inferred from another NP which is semantically associated with it.
Building of previous work on demonstrative expressions, I will propose here treating the three Spanish demonstratives as anaphoric presupposition triggers (Kripke 1990; Van der Sandt 1992) that require a discourse referent, nominally or propositionally realized, which is anchored at a discourse time prior (past) or overlapping (present) the demonstrative utterance time in the case of demonstratives aquel/aquello and este/esto, respectively.On the other hand, demonstrative ese/eso would not require such temporal constraint.These temporal conditions will be of a presuppositional nature and will complete the set of basic and shared presuppositional content for all three demonstratives.In no way does this characterization intend to be a general characterization for demonstratives in all their possible uses (i.e.spatial deixis proper and all their uses as demonstrative anaphors) but only as a way of representing the basic semantic and pragmatic properties of these elements when time contributes to licensing their use in discourse.
The characterization of the Spanish demonstrative pronouns is given in Figures 10-12 in the form of presuppositional Discourse Representation Structures.Note that presuppositional DRSs are built in the same way as regular DRSs (see § 5), but their conditions are pre-conditions of a presuppositional nature.Some of said presuppositional pre-conditions are common to all three demonstratives, namely, Utterance(u) and Agent(x) represent the utterance of the demonstrative expression and the utterer of the demonstrative expression, respectively.The condition Time(u, t) expresses that the utterance of the demonstrative takes place at some particular time (t).The condition Refer in Discourse expresses the idea that the demonstrative is a referring expression that is used by the agent to refer in discourse to a discourse referent, either anaphorically or discourse deictically.Note that no pointing condition has been included in my characterization since in discourse deixis, where reference is made intra-textually, an overt act of pointing (i.e. an ostension) is not necessary.The pre-condition Discourse_Referent(B) refers to the discourse referent to which the agent refers with the use of the demonstrative.This presupposition would be the equivalent to Robert's existence presupposition (i.e.there must be a referent in the discourse so the agent/speaker may use the demonstrative to refer to it.)The referent may denote a concrete entity, an event, fact, proposition or some other discourse referent of higher-order type.The discourse referent can be realized in the syntax either as a noun phrase or with a clause or group of clauses.The conditions Unique(B), Familiar(B) and Activated(B) are specific pre-conditions on the discourse referent.Thus, Unique(B) expresses the idea of informational uniqueness (Roberts 2003).As we mentioned, demonstratives are definite expressions that presuppose the existence of a discourse referent in the context, which is the only element in the domain that is entailed by the common ground to bear the referent's descriptive content.Familiar(B) expresses the idea that the discourse referent belongs in the set of discourse referents that are familiar to the agent either because it has been previously introduced in the discourse, or is contextually entailed or can be inferred.The condition Activated(B) makes explicit the cognitive status of the discourse referent (Gundel et al. 1993). 18 There is an extra time (t') in the universe of the DRSs for demonstratives este/esto and aquel/aquello.This time is necessary to account for the temporal restriction showed by these two demonstratives in discourse.The pre-conditions for the proper use of the distal demonstrative are shown in the presuppositional DRS in Figure 12.The distal demonstrative aquel/aquello would be felicitous in contexts where there is a discourse entity (an event, situation, fact, etc.) of which it is satisfied that said discourse entity is located at time t' (Loc(B): t') and t' is prior to the utterance time of the demonstrative t (t'≺ t); therefore the event, fact or situation denoted by B is to be interpreted as lying at some time in the past of the utterance time.In other words, the discourse referent of demonstrative aquel/aquello is temporally located in the past of the time of utterance of the demonstrative, which is why aquel/aquello shows a clearly marked preference to refer to past events and to be combined or co-appear in discourse with past tenses in line with the results of our corpus study.
The pre-conditions for the proper use of the proximate demonstrative este/esto are shown in the presuppositional DRS in Figure 10.The proximal demonstrative shows a strong preference to co-occur with present tenses in the discourse so this demonstrative would be felicitous in contexts where there is a discourse entity (a concrete entity, an event, situation, fact, etc.) of which it is satisfied that said discourse entity is located at time t' (Loc(B): t') and t' overlaps the utterance time of the demonstrative t (t'⊓ t).In other words, the referent of the demonstrative este/esto overlaps the time of utterance of the demonstrative (present).
Finally, the presuppositional DRS for medial demonstrative ese/eso is given in Figure 11.As we mentioned, our corpus study has shown that the medial 18 These authors propose that different referring expressions have the ability to mark the cognitive status of their antecedents (or they co-referents).Cognitive status refers to the status that a particular referent has in the memory of the interlocutors and this concept finds many similarities with related concepts like salience or accessibility.Gundel et al. (1993) proposed a hierarchy of cognitive statuses (i.e.The Givenness Hierarchy) where the typical status for referents of demonstratives is 'activated', whereas the cognitive status for referents of unstressed pronouns is 'in focus'.
demonstrative does not show any temporal restrictions in its discourse use, hence its corresponding DRS does not include any specific temporal presuppositional preconditions.

Conclusions
A close relationship between demonstratives and tense in Spanish had already been observed a few decades ago but no systematic analysis of this phenomenon had been provided or had it been formalized to date.A detailed study of natural linguistic data from a corpus of Spanish appears to indicate that the connection tense-demonstration is one of a pragmatic nature according to which linguistic expressions that encode or denote time license the use of demonstratives under certain contextual conditions.Generally speaking, these linguistic expressions contribute to setting up adequate temporal frames of reference for a felicitous use of demonstratives.Although grammatical tense appears to be the principal type of licensing expression other linguistic elements (i.e.adverbials, noun phrases, etc.) and even world knowledge also play an important role in establishing the reference time of the discourse.Therefore, the alleged relationship tense-demonstrations should be better conceived of as a relationship time-demonstration.A crucial factor lies in that the aforementioned connection time-demonstration only occurs when speakers employ demonstratives to refer to entities within the textual world (i.e.discourse referents) hence deixis proper (i.e. when physical distance plays a central role in the use of demonstratives) does not share the same temporal constraints.
In this paper, I have analyzed the referring behavior of the three Spanish demonstratives on a corpus of Spanish with an aim to check whether the relationship time-demonstration will find empirical support in Iberian Spanish.The frequencies obtained appear to indicate that distal aquel/aquello does closely interact with tense showing a strong preference to co-appear with past tenses in discourse.This constraint is particularly relevant when the demonstrative is used endophorically, that is, when reference is made within the text to some textual discourse referent.On the other hand, the data from the corpus indicates that the proximate demonstrative este/esto also shows a preference to co-appear with present tenses in discourse in endophoric uses as well.Finally, the medial demonstrative ese/eso exhibits a more heterogeneous behavior.Overall, the medial demonstrative shows a preference to coappear with present tenses but a stronger preference for past tenses than proximate este/esto.Also, it appears that the medial demonstrative is more prone to be combined with tenses that fall outside the categories of present or past, such as the conditional and the future tense.I argue that the relationship time-demonstration is a discourse deictic/anaphoric phenomenon in that speakers do not directly use the demonstrative expression to refer to a time but rather to some discourse entity that is to be interpreted relative to some specific temporal frame of reference (or temporal discourse contexts).It is the specific time of such frame of reference (present or past) that constraints the use of certain demonstratives.Thus, it appears that present time frames of reference strongly disfavor the use of the distal demonstrative aquel/aquello.As regards the theoretical status of the temporal frames of reference, they could also be conceived of as discourse times (i.e. the time at which a fragment of discourse is to be interpreted).Such time, grammatically realized in discourse via sequence of tenses, temporal adverbials, or any other element that may denote time, would act as the necessary licensor of the demonstrative.Given that most of the examples of demonstratives analyzed appear in narrative discourses I have proposed an analysis of the use of distal demonstrative aquel/aquello within the framework of Temporal Discourse Models that allows us to better observe the close relationship between past time and the demonstrative in narrative contexts.
Finally, I have proposed a representation for the three demonstratives constrained by temporal frames of reference as anaphoric presupposition triggers that require that some relevant information can be retrieved from previous discourse for their full interpretation or pragmatic felicity.Thus, as definite descriptions, demonstratives are presupposition triggers as they presuppose that their referent is unique and existent and they are anaphoric (or discourse deictic) for they require, at least, a discourse referent that has been mentioned in the previous discourse and, in the case of temporally marked aquel/aquello, that a past reference frame be established in prior discourse via past tenses, time adverbials or any other expression that denotes a past time.
made his own Christmas presents this year.b.First he dried a bunch of tomatoes in the oven.c.Then he made a booklet of recipes that use dried tomatoes.d.He scanned in the recipes from his gourmet magazines.e.He gave these gifts to his family.

Table 1 .
Spanish demonstrative determiners and pronouns

( 8 )
El equipo de la República Checa fue el vencedor de la XVI edición del Camel Trophy, disputado por tierras de la civilización Maya, en Centroamérica.El equipo español, compuesto por el gerundense Lluís Moret y la manresana Belén Sánchez, terminó en la sexta posición entre veinte equipos.Esta clasificación es la mejor que ha logrado un equipo mixto en este evento.'The team from the Czech Republic was the winner of the 16th edition of the Camel Trophy, celebrated in the land of the Mayan civilization, in Central America.The Spanish team consisting of Lluis Moret from Gerona and Belen Sanchez from Manresa ended up in the sixth position out of twenty teams.This position is the best ever achieved by a mixed team in this event.' (La Vanguardia, 16/07/1995) (9) Sucede, sin embargo, que se puede empatar ante el Mérida pero jamás se puede perder ante el Madrid.Cuando esto sucede las miradas se dirigen casi siempre al palco presidencial.'Nevertheless, it happens that you can end up drawing a tie with Mérida but you can never ever lose against Madrid.When this happens, all the looks usually target the skybox.' (La Vanguardia, 1995) (10) Ya sé que hubo un tiempo en que no nos llevábamos bien usted y yo -añadió-.Pero eso ha pasado a la historia.'I know there was a time when you and I used to get on well -he added-.But that is now history.' (La Verdad sobre el Caso Savolta, 1975) (11) Una vez demostrado por la genética que aquello ocurrió de forma tan sorprendente, cabe esperar la confirmación fósil.'Once shown by Genetics that that happened in such a surprising way, we must wait for fossil confirmation.'(Los Misterios de la Evolución Humana, 2001)

Table 3 .
Demonstrative pronouns in the CREA corpus(Iberian Spanish)

Table 4 .
Demonstrative determiners in the CREA corpus(Iberian Spanish)

Table 7 .
Verbs and Tenses used in the Corpus Study

Table 10 .
Total observed distribution of demonstratives and tense

Table 11 .
Relative frequencies of demonstrative pronouns and tense Figure 1.Relative frequencies of demonstrative pronouns and tense

Table 12 .
Distribution of demonstrative determiners and tense

Table 13 .
Relative frequencies of demonstrative determiners and tense.
Figure 2. Relative frequencies of demonstrative determiners and tense

Table 14 .
A summary of demonstratives in time deixis