1.1. Demonstrative variation in German

Demonstratives such as this and that are among the most frequently used words in language (Brysbaert & New, Reference Brysbaert and New2009). Research on their use in spoken communication has confirmed that demonstratives often play a pivotal role in establishing joint attention between interlocutors on a given referent (Diessel & Coventry, Reference Diessel and Coventry2020; Jara-Ettinger & Rubio-Fernandez, Reference Jara-Ettinger and Rubio-Fernandez2024; Levinson et al., Reference Levinson, Cutfield, Dunn, Enfield, Meira and Wilkins2018; Peeters et al., Reference Peeters, Krahmer and Maes2021; Skilton, Reference Skilton2024). Whether speakers select a proximal demonstrative (such as this) or a distal demonstrative (such as that), across languages, indeed often depends on whether a referent is already in the psychologically shared space between speaker and addressee (Peeters et al., Reference Peeters, Hagoort and Özyürek2015; Rocca et al., Reference Rocca, Wallentin, Vesper and Tylén2019; Rubio-Fernandez et al., Reference Rubio-Fernandez, Long and Özyürek2026). Similarly, when it comes to written language, it has been proposed that the degree of (presumed) accessibility of a referent in the mind of the reader would drive a writer to use either this or that in a given sentence (Ariel, Reference Ariel1990). Recent research, however, interestingly suggests that the discourse genre of a text is the main predictor of writers’ demonstrative choice, at least in Dutch, English and Mandarin Chinese (Maes et al., Reference Maes, Krahmer and Peeters2022a; Peeters et al., Reference Peeters, Schuurman, Zhai, Krahmer, Gu and Maes2025).

Here we further investigate this recent claim in German. There are two main reasons for why a focus on German in this domain can be considered theoretically relevant and important. First, there is a long tradition of studying writers’ demonstrative use in German, with mixed outcomes that have previously been theoretically difficult to integrate, as we will show below. Second, there exists an interesting asymmetry in the demonstrative system used in German written discourse. Like many languages, written German has two frequently used demonstrative variants, as proximalFootnote ¹ demonstrative pronouns and determiners [diese (Frau), dieser (Mann), dieses (Kind); ‘this woman, this man, this child’] and adverbs (hier; ‘here’) can be distinguished from distal demonstrative pronouns (die, der, das; ‘that’) and adverbs (da, dort; ‘there’). However, written German is deprived of unambiguous distal determiners (e.g., die Frau, der Mann, das Kind) as written language lacks word stress to clearly indicate whether a definite article or demonstrative is intended. This asymmetry raises the question to what extent the proximal variant has taken over the function of a distal demonstrative, as has been claimed to happen for recognitional demonstratives, traditionally associated with distal demonstratives (e.g., in English, ‘Do you remember that guy we met in Paris’), but in German suggested to be taken over by the proximal term (e.g., Himmelmann Reference Himmelmann and Fox1996; Schumacher & von Heusinger, Reference Schumacher, von Heusinger, Riesberg, Reinöhl and Hellwig2026).

The variation between German proximal ( diese , dieser , dieses ) and distal ( die , der , das ) demonstrative pronouns has received detailed attention in a number of recent empirical studies (Bader et al., Reference Bader, Portele, Schäfer, Hörnig, von Wietersheim, Konietzko and Featherston2022; Bader & Portele, Reference Bader and Portele2019; Fuchs & Schumacher, Reference Fuchs and Schumacher2020; Patil et al., Reference Patil, Bosch and Hinterwimmer2020, Reference Patil, Hinterwimmer and Schumacher2023; Patterson et al., Reference Patterson, Schumacher, Nicenboim, Hagen and Kehler2022, Reference Patterson, Patil, Ventura, Lialiou, Schumacher and Hinterwimmer2025; Patterson & Schumacher, Reference Patterson and Schumacher2021; Schumacher et al., Reference Schumacher, Patterson, Repp, Gianollo, Jędrzejowski and Lindemann2022). In these studies, differences in use between the two types of demonstrative pronouns have been investigated both on the level of local discourse variables affecting the presumed accessibility and prominence of referents (order of mention: e.g., last mentioned vs mentioned earlier; syntactic function: e.g., subject vs object; thematic role: e.g., stimulus vs experiencer) and on the more global discourse level (e.g., the discourse modality or register at hand).

Compelling evidence for an effect of local variables is scarce. For example, proximal demonstrative pronouns, more so than distal ones, have been suggested to prefer the last-mentioned noun phrase (NP) in the previous sentence as their antecedent (Zifonun et al., Reference Zifonun, Hoffmann and Strecker1997 in Patterson & Schumacher, Reference Patterson and Schumacher2021). However, participants’ use of the two German demonstrative variants showed largely the same pattern in this regard in rating experiments using ditransitive contexts with three antecedents differing in prominence, as in example (1) (Patterson & Schumacher, Reference Patterson and Schumacher2021), and also in sentence completion experiments using sentences such as in example (2) (Bader et al., Reference Bader, Portele, Schäfer, Hörnig, von Wietersheim, Konietzko and Featherston2022). The latter experiments also showed the same behavior in situations in which this last-mention preference was overruled by thematic role preferences in implicit causality sentences. There is indeed growing agreement that der and dieser are quite similar in terms of local structural and semantic variables (Bader et al., Reference Bader, Portele, Schäfer, Hörnig, von Wietersheim, Konietzko and Featherston2022; Patil et al., Reference Patil, Hinterwimmer and Schumacher2023).

1. (1) Die Eigentümerin vermietete die Parkfläche dem Anwohner. Der/Dieser war froh über den langfristigen Vertrag. (Patterson & Schumacher, Reference Patterson and Schumacher2021, p. 1447)

The (female) owner rented the parking space to the (male) resident. ‘That’/‘This’ was delighted about the long-term contract.

1. (2) Der Fischer beeindruckt Peter, weil der/dieser …. (Bader et al., Reference Bader, Portele, Schäfer, Hörnig, von Wietersheim, Konietzko and Featherston2022, p. 408)

The fisherman impresses Peter, because ‘that’/‘this’ ….

Fuchs and Schumacher (Reference Fuchs and Schumacher2020) suggested an interesting alternative account for explaining the difference between German proximal and distal demonstrative use, i.e., their differential capacity to bring about a durable topic shift in discourse. In their sentence continuation experiment, they asked participants to write six sentences following introductory sentences such as in example (3), starting with either an ambiguous personal ( er/sie ), distal ( der/die ) or proximal ( dies-er/−e ) pronoun. The results first show that both proximal and distal variants were predominantly used to establish a topic shift from the subject (‘der Pfleger’) to the object referent (‘den Heimbewohner’). As such, they both obey to the same degree the rule of avoiding to refer to the most prominent referent (‘der Pfleger’). However, the results also indicate that distal demonstratives resulted in more robust topic shifts than proximal demonstratives, as the number of re-mentions of the referent was higher after having started with a distal (vs a proximal) demonstrative. Nevertheless, subsequent research could not confirm this finding in a forced-choice judgment task, as respondents did not more often associate distal (as compared to proximal) demonstratives with more durable topic shift contexts (Schumacher et al., Reference Schumacher, Patterson, Repp, Gianollo, Jędrzejowski and Lindemann2022).

1. (3) Jeden Morgen hat der Pfleger den Heimbewohner gekämmt. Dabei hat er/der/dieser … (Fuchs & Schumacher, Reference Fuchs and Schumacher2020, p. 195)

Every morning, the (male) nurse combed the (male) resident. At the same time ‘he’/‘that’/‘this’ …

On a more global discourse level, it has been suggested that German proximal demonstratives are more commonly associated with the written modality, while distal demonstratives would be more strongly preferred in spoken discourse. For example, in their experiments on the difference between personal and demonstrative pronouns, Patterson et al. (Reference Patterson, Schumacher, Nicenboim, Hagen and Kehler2022, p. 5) decided to use the proximal rather than the distal demonstrative variant because they considered dieser ‘better suited to a written experiment than der , which is perceived by some speakers to be slightly pejorative and is more appropriate in spoken, possibly less formal, contexts’. However, the empirical evidence for this idea is mixed, as witnessed, for example, in Patterson et al.’s (Reference Patterson, Patil, Ventura, Lialiou, Schumacher and Hinterwimmer2025) spoken version replications of two earlier, written experiments (Hinterwimmer & Patil, Reference Hinterwimmer and Patil2022; Patterson & Schumacher, Reference Patterson and Schumacher2021). In their first replication experiment, no difference was found in participants’ acceptability ratings between the written and spoken use of proximal and distal demonstratives. In their second experiment, using rather unusual experimental sentences with two demonstrative NPs as in example (4), the use of two distal demonstratives (‘Die … dem’) turned out to be more acceptable in the spoken as compared to the written version. As a tentative suggestion, Patterson et al. (Reference Patterson, Patil, Ventura, Lialiou, Schumacher and Hinterwimmer2025) propose that both modality (spoken vs written) and register (neutral vs informal) played a role here. Specifically, they conjecture that the use of two distal demonstratives may have pushed the spoken sentence from a neutral into a more informal register, while the written equivalent with two distal demonstratives would feel less acceptable because of the unnatural combination of the written modality and informal language use.

1. (4) Im Konzert gestern Abend hat sich eine Musikerin mit einem Stimmnachbarn gestritten. Die/Diese hat nach dem Konzert kein Wort mehr mit dem/diesem gesprochen. (Patterson et al., Reference Patterson, Patil, Ventura, Lialiou, Schumacher and Hinterwimmer2025, pp. 25–26)

In the concert last night, a (female) musician argued with a (male) colleague. ‘That’/‘This’ did not speak a word anymore with ‘that’/‘this’ after the concert.

This explanation seems in line with experimental evidence for the role of register in German demonstrative use. Concretely, Patil et al. (Reference Patil, Bosch and Hinterwimmer2020) showed a clear preference in forced-choice tasks for distal demonstrative pronouns in informal contexts and proximal pronouns in formal contexts. However, as these results were obtained on the basis of a small set of similar, constructed examples only, the scope of these ‘global’ register preferences remains unclear, for example, in view of corpus studies finding relatively large numbers of distal pronouns in formal conversations (Weinert, Reference Weinert and Weinert2007) and newspaper texts (Bosch et al., Reference Bosch, Rozario and Zhao2003). As such, it remains to be seen whether the experimental effects of register hold under more natural circumstances.

In sum, on the basis of existing studies, it seems fair to conclude that the influence of local variables on the German writer’s use of proximal vs distal demonstratives is limited, while global variables such as modality and register may tentatively be considered to play a more important role in explaining which demonstrative a writer uses in a given sentence or text. Yet, the picture sketched by the existing literature remains incomplete. First, the studies discussed above almost exclusively focused on demonstrative pronouns referring to animate referents. It remains to be seen whether existing findings generalize to demonstrative NPs and adverbs, and to demonstratives referring to non-animate referents and non-nominal (i.e., abstract, propositional) antecedents (cf. Maes et al., Reference Maes, Krahmer and Peeters2022b). Second, the role of global discourse variables such as modality and register deserves to be investigated in a more systematic way, and not simply as a by-product of the study of local variables. In other words, local and global variables should be contrasted within the same study and taken into account in the same research design. Third, almost all previous experimental work on demonstratives in German circumvents natural written language production. Most of the theoretical proposals put forward are tested through constructed sentences that are implemented in forced-choice, continuation or rating tasks, where participants are invited to play the part of language expert or evaluator, fully aware of the research focus on these little demonstrative words, while at the same time largely kept in the dark about the type of discourse the sentences would fit in. The writers’ behavior in such environments is likely quite different from their behavior in writing full texts in a specific genre given a particular discourse goal. In particular, the role of higher-order discourse variables can hardly be investigated reliably in such restricted experimental contexts. And we know that participants’ intuitions on demonstratives not necessarily match the reality of their everyday use (e.g., Piwek et al., Reference Piwek, Beun and Cremers2008).

Finally, even when suggestions on the difference between proximal and distal demonstratives in German are plausible and some results replicate, it remains unclear how to explain them in an overarching theoretical framework. Why would proximal demonstratives be strongly associated with writing and formal language, and distal demonstratives with speaking and informal language (Patterson et al., Reference Patterson, Patil, Ventura, Lialiou, Schumacher and Hinterwimmer2025)? To advance our understanding, an experiment is required that is based in an established theoretical framework and allows for systematically analyzing the influence of global vs local variables in German writers’ naturalistic written demonstrative use.

1.2. The present study

In the present study, we aim at providing a comprehensive account of the difference between proximal and distal demonstrative use in German. Our theoretical starting point is the conceptual framework set out in Peeters et al. (Reference Peeters, Krahmer and Maes2021). This theoretical account considers demonstrative variance to be dependent on a combination of physical, psychological and referent-intrinsic factors. The relative importance of these factors is influenced by sociocultural parameters, such as personal characteristics of the individual language user and affordances offered by the physical or discourse context. For demonstratives in written discourse, the framework considers the affordances provided by genre experiences as most crucial in explaining demonstrative variation.

Discourse genres are commonly defined in terms of pre-existing assumptions on how writers interact with addressees and referents (Biber, Reference Biber1993; Weaver & Kintsch, Reference Weaver, Kintsch, Barr, Kamil, Mosenthal and Pearson1991). These assumptions are firstly related to different discourse goals, for example, the goal of the writer to narrate, expose or evaluate (Mar et al., Reference Mar, Li, Nguyen and Ta2021). Lifelong experiences of writing and reading texts given specific goals and intentions likely lead to differences in how language users conceptualize how they use and perceive language across genres. Specifically, demonstratives are considered as ways of expressing default interactional assumptions connected to particular discourse genres. Concretely, proximal demonstratives are seen as expressing the primary focus of a writer on the topic of a text, thus stressing in the first place the tight connection between a writer and the topic they write about. Distal demonstratives, on the other hand, primarily express the interaction of writers with their readers, as they provide discourse referents with a flavor of familiarity (Peeters et al., Reference Peeters, Krahmer and Maes2021). As such, it is proposed that an increasing preference for distal demonstrative anaphors should be observed when the role of the addressee becomes more prominent in the discourse setting at hand (as in narrative discourse, for instance), while an increasing preference for proximal demonstratives should be observed when writers are focused more on the topic of their text, as in an expository context (Peeters et al., Reference Peeters, Krahmer and Maes2021).

Thus far, both corpus research in English (Maes et al., Reference Maes, Krahmer and Peeters2022a) and text production experiments in Dutch and Mandarin (Peeters et al., Reference Peeters, Schuurman, Zhai, Krahmer, Gu and Maes2025) have confirmed these preferences in writers’ demonstrative use by looking at the presence of demonstratives across different genres. Obviously, the genre differences hypothesized here are related to notions like modality and register prominent in the existing German demonstrative literature. It is likely to assume that the narrative genre can be strongly associated with informal spoken and written discourse settings and the expository genre with more formal writing. The conceptual framework outlined above suggests that, underlying these notions, however, one should situate the assumptions related to the differential relations between writer, the referents discussed in their text and their addressee(s) across discourse genres.

Methodologically, in the current study, we replicated the established setup that was used in Peeters et al. (Reference Peeters, Schuurman, Zhai, Krahmer, Gu and Maes2025) in the study of written demonstrative variation in Dutch and Mandarin. As such, we here asked German respondents to write two short texts in different genres (narrative and expository) based on two different scenarios. This setup allows for contrasting a potential influence of discourse genre with the effect of ‘local’ variables such as the relative distance between demonstrative and antecedent, and whether the demonstrative is in subject or initial position in the sentence or not. In line with the conceptual framework discussed above, we expected to find an overall preference for proximal demonstratives in expository texts, and for distal demonstratives in narrative texts (Peeters et al., Reference Peeters, Krahmer and Maes2021). Under such an account, any meaningful remaining variation in demonstrative use might be attributed to local discourse variables (Maes et al., Reference Maes, Krahmer and Peeters2022a). In contrast, if written discourse is strongly associated with proximal demonstrative use across the board in German (Patterson et al., Reference Patterson, Patil, Ventura, Lialiou, Schumacher and Hinterwimmer2025), we should observe substantial proximal demonstrative usage regardless of genre.