The semantics of space and time of the far right. A neopragmatic study of landscape interpretation by generative Al

Abstract

Mediatisation, digitisation and algorithmisation irritate society. New possibilities of communication using AI potentiate forms of self-description and self-observation and serve, among other things, to influence politics. In this study, we investigate the question of the extent to which landscape-related visualisations of the extreme right can be accessed via generative AI and how semantics of space and time are constructed through data. The case study is based on an exploratory study (Frankenberger et al., 2024) and visualises landscape based on the electoral program of a right-wing populist German party in 2021. While a closed image of landscape and society is computed, the visualisation of a neutral prompt shows an open development. The results are analysed in terms of syntax, semantics and pragmatics and interpreted from a neopragmatist and systems theory perspective. It becomes apparent that the computation of semantics of space and time is contingent and reduces complexity and that it is required to access these reflexively.

Headnote

Abstract: Mediatisation, digitisation and algorithmisation irritate society. New possibilities of communication using AI potentiate forms of self-description and self-observation and serve, among other things, to influence politics. In this study, we investigate the question of the extent to which landscape-related visualisations of the extreme right can be accessed via generative AI and how semantics of space and time are constructed through data. The case study is based on an exploratory study (Frankenberger et al., 2024) and visualises landscape based on the electoral program of a right-wing populist German party in 2021. While a closed image of landscape and society is computed, the visualisation of a neutral prompt shows an open development. The results are analysed in terms of syntax, semantics and pragmatics and interpreted from a neopragmatist and systems theory perspective. It becomes apparent that the computation of semantics of space and time is contingent and reduces complexity and that it is required to access these reflexively.

Keywords: semantics of space, semantics of time, artificial intelligence, generative ai, landscape, neopragmatism, systems theory.

Article Info: Received: February 13, 2025; Revised: April 12, 2025; Accepted: April 23, 2025; Online: May 31, 2025.

Introduction

Comparable to earlier mediatisation processes (Krotz, 2012, 2015; Lundby, 2014), such as the varying forms of communication that emerged after the introduction of writing or letterpress printing, digitisation (Baecker, 2018; Nassehi, 2019) and algorithmisation (Beer, 2017; Esposito, 2022) also irritate the normal functioning of society (Luhmann, 2016 [1999], pp. 55-100). Irritation in the sense proposed here refers to the following, based on Luhmann's (1989 [1980], 2018 [1998], p. 118) social theory: Communication systems are formed in dependence on the available semantics and, through all communication processes, create the basis for the semantics that can be used in the future. This means that at every moment of communication, we can only draw on the available knowledge, interpret it and recombine it. In Rorty's words, we can therefore create new descriptions on the basis of oral communication, texts and literature, images and videos or all products of art. However, if new technologies suddenly provide us with an increased supply of new knowledge, we are irritated because we first have to organise the abundance of external additional information internally. This means that the structure of society is constantly changing, as we can access and combine a wide range of information using new technologies. With increasing digitisation and the spread and application of artificial intelligence, we are being shown how profound such a change can be. The two processes not only dynamise communication processes and detach them from the occupation of places in a space imagined as material (Luhmann, 2018 [1998], pp. 302-315; introductory to Luhmann: Arnoldi, 2001). Rather, both the dynamisation itself and the structures of society that have to prove themselves in this way can be observed. Consequently, it is possible to provoke and observe the retrieval of specific semantics that depict spatial and temporal references - and to gain insights into ways of dealing with complexity and contingency.

While geography is already beginning to tap into this with research fields in digital and visual geography, media geography and the analysis of spatial semantics (e.g. Bork-Hiiffer et al., 2021; Doring & Thielmann, 2009; Redepenning, 2006; Schlottmann & Miggelbrink, 2015), we see potential in triangulating theories from a neopragmatist perspective (Kiihne, 2024b, pp. 162-164) in order to supplement research into materiality with research into virtuality. This is not to say that geography, sociology or philosophy have not yet made any progress in this regard. However, an opening of geography for the exploration of completely abstract spaces and spatiality (in the digital) did not result, even though the potential and challenges of this have been identified, for example by Lippuner (2007, pp. 180-184). With this article, we would like to offer a first suggestion for observing this abstraction, but at the same time - in the tradition of pragmatism - maintain a reference to the everyday world. In this respect, we are analysing the spatial category of landscape, which is present in everyday life, following Miiller-Salo (2020). The spatial perspective of extreme right-wing thinking is of interest not only to academics, but also to a wider public, because it deliberately refers to material space. In this respect, it not only follows on historically from conservative metropolitan criticism, but also explicitly refers to the preference for the rural, both in terms of material spaces and 'rural-traditional' ways of life. This preference was normatively charged with the aim of 'ruralising' large cities (Franke, 2019; Ipsen, 1992; Korner, 2006; Kiihne, 2019).

Landscape, regardless of whether rural, urban or suburban, is constructed through conventions, through multisensory processes of perception and their connection to proven vocabularies, and is reproduced in daily communication (i.e. Hokema, 2013; Micheel, 2012). Consequently, contingent descriptions of spatial sections are constructed and then categorised as landscape - in one way or another (Lammchen, 2023). The concept of inverse landscapes makes it possible to address these construction processes (see 2). Furthermore, this contingent approach to complexity leads to a self-description that is dependent on space and time. By fixing observations in data form (texts, images, audio, etc.), we create social versions of current world observation, whereby existing patterns are reproduced (narratives, myths, etc.), but new forms of observation (fiction, utopia/dystopia, etc.) are also offered and vocabularies can be further developed. Through this form of self-description, we also offer forms of self-observation, for example by entering prompts into generative AI. Consequently, it is possible to view new semantics of space and time at the push of a button (Lammchen, 2024a, 2025); however, we are also challenged to approach this phenomenon reflexively, that is, to fundamentally recognise experiences of contingency and thus avoid instrumentalising essentialist truths as disruptions of contingency (see 3). From this perspective, we formulate the following thesis: semantics related to space and time can be accessed via a syntax (as data sets that can be related by algorithms) and, if the reflection on the contingent calculation of the world - in this case, landscape - is lacking, can develop imperative references (pragmatics). The visualisation of selected vocabularies from the 2021 party programme of the German right-wing populist party 'Alternative fur Deutschland' (AfD) serves as an empirical example and is based on an exploratory study (Frankenberger et al., 2024). These visualisations are compared with neutral visualisations, which, however, also remain contingent. To interpret the results, we develop theory-based tools for observing syntax, semantics and pragmatics (Kiihne et al., 2025) (see 4 and 5). The selection and approach are based on the increasing use of Al-generated images of extreme right-wing parties (an example of this is the reception of the election posters of the 'Biindnis Zukunft Hildburghausen' in the local elections in Thuringia in 2024). Finally, we discuss the results in the context of our thesis and with regard to the challenges of a liberal, solidary society in dealing with contingent forms of self-description and self-observation through (artificial) visualisations of semantics of space and time.

An approach to neopragmatist landscape research

The philosophy of neopragmatism is based - in a nutshell - on the tradition of classical US-American pragmatism and continental European philosophy of language (Rorty, 2023). From the former, among other things, the preference of thinking arises from the consequences of action and practical utility, and from the latter, the understanding of the creation of the world in vocabularies. Vocabularies can be understood as interrelated statements that form a system of meaning that differs from other vocabularies. Vocabularies have arisen from other vocabularies, they have prevailed over other vocabularies and in turn will (most likely) be replaced by other vocabularies (Rorty, 1997 [1989], 1998). The emergence of differentiated vocabularies can be understood as a phenomenon of linguistic division of labour (Putnam, 1990, p. 43). As a result of the increase in complexity within language (and society) on the one hand, and the limited number of relations between expressions of linked criteria (Putnam, 1990, p. 43) on the other, this necessity of a division of labour through the differentiation of vocabularies arises. According to Luhmann (1984, 1997), this differentiation can be understood as a complexity reduction, i.e. the selection of relations between elements.

The differentiation of vocabularies and their antagonistic position to each other is common to discourse theories (i.e. Foucault, 1996, 2007 [frz. Original 1971]; Laclau & Mouffe, 1985). What distinguishes Rorty's theory of vocabulary from this is the approach of redescription, which arises from modifying our practices so that new descriptions of what has happened so far are taken into account (Rorty, 2003, p. 297). Redescription does not mean replacing an old vocabulary with a new one, but integrating it in whole or in part into a new vocabulary (Topper, 1995). In recourse to Dewey, it is about reconciling older innovations with newer ones (Rorty, 2023, p. 252).

Contingency, as the non-necessary but also the non-impossible, has a central significance for Richard Rorty's philosophy (Rorty, 1998, 2009 [1989]): He refers to the contingency of language, the self and society. Language is - as has become clear above -subject to constant change, without having a nature or a goal. The self, too, has no essential identity, but constantly reimagines itself within contingent boundaries. He also rejects the idea of universal or necessary social structures. As products of social developments, societies are changeable - and liberal democracy is not a teleological endpoint of social development either, but must be defended against authoritarian tendencies constantly (Rorty, 2009 [1989]. Accordingly, the basis for a society is not the nature of a people, but a solidarity based on compassion (Lohmann et al., 2024; Rorty, 2009 [1989]). Just as there is no thing in itself, there is no such thing as the one correct description of a thing: There are only descriptions that, by relating the thing to other things and embedding it in contexts, serve our current, diverse needs (Rorty, 2023, p. 54). This highlights the panrelationalist view of neopragmatism, according to which something only comes into being in relation to other things (Rorty, 2023; more detailed introductions to Rorty's thinking can be found in: Calder, 2003; Horster, 1991; M. Miiller, 2021, 2023; Reese-Schafer, 2016).

If we approach landscape from this neopragmatist perspective, it is constitutively bound to vocabularies (Kiihne, 2024b). Landscape thus arises from linguistic relations that are then projected onto materialities or virtuality (Kiihne, 2021; Kiihne & Edler, 2022; Kiihne et al., 2021). Landscape-related vocabularies are - like the other vocabularies -characterised by changeability. This becomes clear when we consider the historical development of the understanding of landscape, which has shaped a large semantic field (Hard, 1969) that includes material, moral, aesthetic, normative, etc. aspects (for more details see, among others: Berr & Kiihne, 2020; Hokema, 2013; G. Miiller, 1977; Schenk, 2013). With semiotics rooted in American pragmatism (Peirce, 1991,1998), Morris (1938) distinguishes between syntactic, semantic and pragmatic dimensions of signs, which - if landscape is understood as the result of vocabulary production, as here - can also be applied to landscape (see for example: Denzer, 2024; Jaworski & Thurlow, 2010).

Syntax refers to the relation between signs, in this case the attribution and inscription of relations, not only in material or virtual space, but also in relation to elements of social and individual construction of landscape. The semantics of landscape refers to the relationship between signs and meanings, such as the attribution of interpretations, categorisations and evaluations in and to spaces designated as landscape (whether material or virtual). The pragmatics of landscape refers to the relations between signs and sign users, i.e. how we interact with spaces that are interpreted as landscape. This can be passive, in that action is controlled, or active, in that the material (or virtual) space interpreted as landscape is intervened in.

From these observations, it is clear that, according to the understanding of landscape represented here, landscape is highly contingent. Landscape is just as contingent in its material/virtual, individual and social dimensions as it is in terms of time. What is interpreted as landscape is therefore neither arbitrary (because there are certain social, culturally bound conventions) nor determined. The concept of inverse landscapes makes it possible to concretise landscape contingency by specifying particular landscapes from the space of contingent landscapes visually, textually, and also acoustically (other sensory approaches are also possible) (for the original concept Kiihne, 2013a, 2013b; in this understanding presented here: Kiihne et al., 2025; Kiihne et al., 2023). Before we turn to the construction of inverse landscapes of the far right, we focus on the development of semantics of space and time, which, not at least, underlie the contingency of landscape.

Semantics of space and time as data-based distinctions

If we considered language as a tool for creating the world in vocabulary above, we will expand this perspective in this section to include the description of processes for creating the world in data. Complementary processes of mediatisation, digitisation and algorithmisation (see 1) require an extension of the linguistic-philosophical observation to include a technical-sociological perspective, which will be developed below based on modern systems theory. This theory is based on the theory of social systems by the German sociologist Niklas Luhmann (1984, 1997) and combines, among other things, cybernetic (Foerster, 1995), mathematical (Spencer Brown, 1969), biological (Maturana & Varela, 1987; Varela et al., 1974), psychological and linguistic theories (Glasersfeld, 1988,1995) of the 20th century. In numerous publications (exemplarisch: Baecker, 2018; Esposito, 2022; Nassehi, 2019), the potential of this theory for observing varying modes of communication and social evolution has already been developed to some extent and also applied to geography (Goeke & Lippuner, 2011; Kiihne, 2024a; Lammchen, 2024a, 2024b, 2025; Redepenning & Wilhelm, 2014). This potential is realised through a comparative transparency and pragmatism, like Rorty's philosophy - it makes its own distinctions the subject. It assumes its own distinctions as contingent forms of describing the world and attempts to (re)observe those own and foreign distinctions from a second-order perspective (Luhmann, 1990, pp. 96-103). If, for example, a person describes a landscape as a normal, familiar part of their homeland, systems theory also emphasises the importance of examining the how of the distinction. Essentialist or objectifying observations are excluded - only the contingent distinction itself, which could also have been different, is of interest.

If we apply these considerations to worlds 0/data-generated images, it is striking that we experience contingency on three levels. Firstly, we have to categorise a mediated image that is supposed to represent a landscape as such (Lammchen, 2023) and evaluate it aesthetically, emotionally or in terms of content (expert knowledge etc.). Secondly, computer-generated landscapes and in particular those created by generative AI (Dall-E, Midjourney etc.) are the result of a contingent linking of text-image pairs. In these cases, both the composition of the training data set and the algorithm-based calculation of supposedly new landscape images remain non-transparent. In this case, the contingent setting of distinctions by data as observer (Nassehi, 2021, pp. 108-113) reveals the visual deviance (Lammchen, 2025) that we still perceive in many cases as such and can distinguish from a real recording.

Thirdly, we experience contingency in our encounters with visualisations and by using the computer itself. That output is the result of inputs made by us and others, which are intertwined in a parasitic and convivial way (Lammchen, 2024a). We encounter ourselves using a computer - a virtual contingency experience that is both confusing and fascinating at the same time (Esposito, 1993). But to what extent can a shift be seen in the three-step syntax-semantics-pragmatics through this development?

As indicated above, the term syntax in semiotics and linguistics encompasses the grammatical rules of how words and sentences are put together. This term is also used in computer science, for example to describe the linguistic patterns recognised and generated by large language models (LLMs). In the case of generative AI models, however, a training data set is not used merely to calculate the next letter of a word or the next word in a sentence using probability models, but also to create sentences and sections with a primitive semantics (O'Grady & Lee, 2023). This means that AI models understand context independently to a limited extent. However, understanding does not mean reflection or charging with meaning - only sequences of negation processes are used to calculate what can and cannot be said and when; along the lines of pattern recognition, the system tests a self-generated uncertainty with observed resistance and thus recursively determines a realisable intrinsic value from moment to moment (Luhmann, 2018 [1998], pp. 1126-1127). However, the algorithm remains at the level of a copy of human communication. The system is not intelligent, but the communication is merely artificial (Esposito, 2022). The same principle applies when calculating new text-image pairs from existing ones. Models understand which combinations prove successful in their contingent composition - similar to the viable generation of the world through language and the linking of language with materialities or virtualities by individuals (Glasersfeld, 2010). In this way, generative AI models generate semantic knowledge bases that we can interpret as such, but the models remain at the level of calculating meaning-processing rules (Luhmann, 1993 [1980], pp. 17-21). Consequently, semantics are expanded to include further possible combinations, and at the pragmatic level, too, new possibilities arise for constructing and controlling (Goeke et al., 2015) landscape-related visualisations, which are thus not free from possibilities of influence. This means that by entering keywords or entire sentences into generative AI models (as so-called prompts), it becomes possible to combine familiar elements by a new pattern and place them in a specific (also political) context. This corresponds to the core of the concept of the inverse construction of landscape (see above) and results in the emergence of new semantics of space and time.

Semantics of space create a sense of the natural, reduce complexity and serve as disruptors of contingency (Redepenning, 2006, p. 131). In the case of text- and image-based visualisations using generative AI, it is possible, for example, to extend notions of what a landscape should be with elements that both reduce and increase complexity - and this regarding the interface between semantics and pragmatics. In this regard, temporal contexts are also unveiled. Following Nassehi (2008, p. 20), we define descriptions of the past, present and future as semantics of time, which in turn vary along the social-structural changes. The future is dynamically constructed through the large amount of produced information (Bateson, 1985, pp. 488-489). This makes it possible to determine semantics of space and time as data-based distinctions - the reception of visualisations and texts generated by generative AI shape our view of space and time.

Based on these observations, the following can be stated: If social conventional notions of landscape are integrated into training data sets of generative AI models via text-image pairs and expanded with prompts that derive from a specific political agenda, then both repositories of knowledge about space and time (for example, in the context of dystopia/utopia) can be identified. Considering that the training data set of Dall-E, on which the following investigation is based, is not free of ideologically charged content, the input of specific prompts (syntax) can provoke primitive semantics. Accordingly, we postulate that political imaginings are reflected in generative AI landscape visualisations. This not only demonstrates the influence of visualisations on concepts of the future, represented here as inverse landscapes, but also highlights the recursive process of contingent distinction, which enables us to act as observers of the observation.

Method of data collection and interpretation

The aim of our empirical analysis is to create visualisations of generative AI using contingent landscape-related vocabularies and to evaluate and interpret them with regard to semantics of space and time. To illustrate how semantics of space and time vary depending on the input (i.e. in relation to the vocabularies), we developed a prompt based on the party programme of a German right-wing populist, in some parts even right-wing extremist party (AfD) from 2021. In the context of an explorative study (Frankenberger et al., 2024), frequencies were calculated for 1,285 terms from the aforementioned 210-page programme. To develop a prompt regarding landscape vocabularies, we only considered the 15 terms that were mentioned most frequently (further conjunctions or articles could be filtered out using a stop list in the MaxQDA application). For seven of the 15 terms, we expected the potential to be able to combine them with the term landscape, which are shown below with the respective rankings (R) and frequencies (n): AfD (R:i; n=252), Germany (R:2; n=i49), German (R:s; 11=77), family (R:7; n=62), child (R:8; n=5i), state (R:9, n=48), land (R:i3; n=45). Terms such as 'not' (R:3; n=i37), 'only' (R:6; n=72) or 'citizen' (R:io; n=47) were not included because they either did not meaningfully supplement the prompt or because they were categorised as intersections of related terms (e.g. citizen and state). The acronym EU (R:i5; n=4i) was not included because the AfD represents an anti-(union)European position, so it also frames land and landscape primarily as a national construct. With this selection, which was necessary to clarify the context, we had a prompt suggested by ChatGPT itself and added the term landscape. We decided to work with OpenAI, in this case ChatGPT 40 in combination with Dall-E 3, because it uses a general and broad training data set with billions of text-image pairs as a basis. We created both the prompt and the visualisations on 10 December 2024. Based on our input 'Create a prompt that combines the following terms in a meaningful way: Landscape, AfD, Germany, Germans, family, child, state, land' ChatGPT 40 generated the following: 'A stylised landscape that is symbolic of Germany: rolling hills, an idyllic village with a German flag waving, a traditional German family with parents and children walking hand in hand. In the background, the silhouette of a state that protects and organises the country can be seen, with a focus on community and cohesion. The style should appear calm and harmonious and emphasise the value of tradition and family'. To include a time reference, we expanded the generated prompt with the following phrases: 'at the end of the 19th century, 'at the end of the 20th century', 'at the end of the 20th century'. We then generated the graphics with Dall-E 3.

To interpret the results, we draw on the distinction between syntax-semantics-pragmatics introduced above. We therefore developed our own theory-based tools of observation to analyse the visualised quantitative data according to the following categories (Kuhneet al., 2025):

Syntax: Landscapes can be described as the result of relations that are synthesised into material spaces on the basis of social conventions. When dealing with the syntactic dimension, we observe whether and to what extent the Al-generated images follow these conventions.

Semantics: The social and individual construction of landscape is based on symbolic charges that have developed over centuries and are socially and culturally quite diverse (Bruns & Munderlein, 2024; Kiihne, 2019). In this part of the analysis, we examine whether and to what extent symbolism, in particular specific extreme right symbolism, is being actualised.

Pragmatics: The individual and social construction contains not only patterns of categorisation and interpretation, but also - based on these - normative aspects. In this respect, the relationship to landscape is often associated with an aspect of instruction. This can be reported as carrying out a certain action (based on elements of the material or virtual space interpreted as landscape), but also - based on normative elements of construction -an intervention in the structures and functions of the material or virtual space interpreted as landscape. In this respect, our study addresses the question of the extent to which a reference to an instruction can be established in the generated landscapes.

Semantics of space and time of the extreme right in the landscape interpretation of artificial intelligence

The following illustration shows the results of the visualisations using the above-mentioned prompt based on the AfD programme with varying time references (Figure l).

At the syntax level, it is noticeable that the Al-generated prompt already embeds the text entered in a context based on the election programme, therefore calculating relations internally. Consequently, the idyllic village embedded in a landscape is already recognised as the origin of the traditional German family with parents and children and is then visualised. Especially in the first two illustrations, the state that protects and governs the land is symbolised by the depiction of an institution (in this case specifically: church and monastery), which is spatially located above the surrounding gentle hills (semantics). At the end of the 21st century, this imagery is fading. The interpretation of the closed networklike structure, which is positioned around the setting sun, remains speculative. It is possible that the landscape is seen as a shelter from an external, increasingly technical and mediatised sphere, which, however, extends into it via the visualised strings, a central topos of romantic landscape construction (see in more detail: Kirchhoff & Trepl, 2009; Kiihne & Berr, 2024; Schenk & Steinkriiger, 2020). In the context of the considerations regarding inverse landscapes, the subject of the image analysis is also what cannot be seen: urban settlements. In the construction by the AI following the electoral programme, a German landscape here is a purely rural landscape and this at all times. This clearly shows the extreme right-wing party's already stated preference for the rural (physically spatial, as well as socio-culturally). A reference to pragmatics would then be the preservation of established and desired forms of life in spatial and temporal dimensions. Order is also achieved through small-scale settlements connected by rarely used paths (regardless of how inefficient this may be) and through economic activities in the primary and secondary sectors (e.g. visualised field work and half-timbered buildings). This normative charging of material space constructed as landscape also dates back to the period of industrialisation and urbanisation. In this period, the (traditional rural) landscape was idealised, especially by people with a higher level of symbolic capital (Bourdieu, 1984).

This idealisation ignored the living conditions of the rural population, who also lacked symbolic capital for the rural view (i.e. Berr & Schenk, 2019; Burckhardt, 2006; Kiihne, 2018).

The stylised landscape is displayed in (late) summer. The incidence of light and the colours used resemble a stylised representation as a portrait. In the transition from the 19th and 20th to the 21st century, the warm and saturated colours change to a cooler, smoother form of representation. The family is depicted from behind in all the graphics - we look from the outside at an inward-pointing and closed community that achieves its cohesion through enduring structures and traditions that are maintained - landscape serves as a symbolic anchor in every time sequence. The images also convey normative observations (family, gender, but also homogeneous phenotype), which can be interpreted as guiding distinctions for preserving this idyll. The image generated here refers back to the idea of a landscape that draws in particular on stereotypes that emerged during the period of social modernisation in the sense of a transition from an agrarian to an industrial society (especially in the 19th century) (Kirchhoff & Trepl, 2009; Kiihne, 2018; Vicenzotti, 2011).

Furthermore, several semantic elements should be highlighted. The prompt developed by ChatGPT syntactically complemented the relevance of including waving flags as a symbol for linking landscape, nation/state and community. Consequently, oversized elements, which are typical for present-day AI systems, are added in all three illustrations, although inappropriately related, in order to visually express the computed relation. During the course of time, however, the flags are shown becoming increasingly smaller, even if the representation of the flags is in neither a subtle nor a sophisticated way. Furthermore, the use of the depicted symbols and objects is remarkable: at the end of the 19th century, the national and merchant flag of Germany was black, white and red, not black, red and gold. The German Bundesadler is added to the first two images; a national symbol that goes back to the beginnings of the Holy Roman Empire; however, by the end of the 19th century, the German Reichsadler was the generally used national symbol. The car in the foreground on the left shows a design from the 1930s (image 3 in Fig. 1). Moreover, wind power - an important political issue for the party - is a central component in all three graphics. However, the technical development is interpreted rather declining. The windmills in the first graphic (end of the 19th century) appear most modern; an American windmill is depicted in the third graphic. In addition to the error rate in AI relations, this point could also be discussed in more detail.

In contrast to this contextualised visualisation, we developed a second neutral prompt and had images generated analogous to the three-time references (Prompt 1: 'Create an image for the following prompt: 'A German landscape at the end of the 19th century"; Figure 2).

On a syntactical level, it should be emphasised that the calculated German landscape once again follows socially evolved conventions but is designed to be symbolically and semantically more neutral. This means that landscape is not stylised according to nationality, but synthesised entirely without symbolism, such as flags. Landscape does not appear to be tied to a family, a homogeneous group with traditions, a closed community, but is simply the result of anthropogenic intervention. People are not depicted in any of the graphics. Of course, elements such as the construction of roads, half-timbered houses or fields can be interpreted as traditional, which are again part of the calculated pattern. However, these are subject to change. This dynamic is harmoniously integrated into a low mountain range landscape - a landscape type that has been calculated as average for Germany. In particular, the graphic from the end of the 19th century is syntactically better than in the first sequence.

Overall, the first graphic shows a small-scale, densely wooded landscape characterised by grassland, with no contrast being created. Fast-growing cities in Germany are not visualised as a counterpoint - Germany apparently developed exclusively in a rural idyll, again characterised by small settlements, religion and handicrafts. At the end of the 20th century, we see very small parcels of land that can hardly be used economically, somewhat unmotivated fence positioning and no such clear presence of wind farms. Here, urban centres are shown with industry, but without a demarcation of the imperative references (pragmatics) and with a smooth blend into the horizon. Industry and city are not seen as contrary, but rather as part of the landscape. In the third graphic (end of the 21st century), the unmotivated road layout is striking and is complemented by the somewhat incomprehensible positioning of wind and solar panels. Futuristic buildings are also scattered around with no visible access routes; the landscape is shown to be more mountainous, and urban centres are integrated into what were once small-scale settlements.

We can already see at this point that landscape is subject to general change and is not exclusively symbolised and stylised through community, family, nation and community (semantics). The increasing industrialisation of the landscape is striking. While industrial elements are excluded from the representation for the end of the 19th century, they are depicted in the background at the end of the 20th century, supplemented by wind turbines in the middle ground.

For the landscape construction at the end of the 21st century, the AI assumes a greater mix of functions, with technical installations for energy generation in particular prevailing over the otherwise - as shown - representation oriented towards ideas of a traditional landscape. This industrialisation of rural areas, interpreted as landscape, would inscribe the landscape dystopia of the Heimatschutz movement since the end of the 19th century (see: Berr, 2019; Korner, 2006; Piechocki, 2010; Sieferle, 1985).

Once again, all three visualisations are shown in (late) summer. The insight into the non-transparent training data set (syntax) could provide information about the reason for this. Again, warm and saturated colours are predominantly used, but the future is not only illustrated in cool colours, but sometimes in the most harmonious way. In the third graphic (end of the 21st century), a small car is depicted, which can be interpreted as representing electric mobility. By contrast, the AfD, a party that campaigns for the continued use of petrol engines, is represented in the third graphic of the first row of images by a tractor and an older car.

From this comparative analysis of the generated image pairs, we conclude the following:

o Along the visualisation of landscape based on the AfD party programme (Fig. 1),...

... landscape is constructed as a closed sphere without a counterpart (city, industry, etc.) at all times, but an external, intangible complexity (network in Figure 3) is contrasted to this idyll.

... landscape is stylised as the anchor point of a homogeneous people (family, phenotype, etc.).

... landscape is depicted as the habitat of a traditional community, making it easier for supporters to derive a imperative reference (protection) (pragmatics).

... a traditional community is associated with traditional economic forms (agriculture and crafts).

... order is embodied by spatially situated institutions (churches, monasteries) and by flags and national symbols.

... future is designed as a challenge to the existence of this normatively charged construction (cooler colours, less rural population etc.).

o Along the visualisation of landscape based on the neutral prompt (Figure 2)

... landscape is designed as an open sphere, which is increasingly penetrated by urban and industrial forms of life and economy.

... landscape is interpreted as an anthropogenic vehicle that is not necessarily bound to a population that can be defined as homogeneous.

... it is therefore not possible to tie a constructed traditional community to landscape.

... consequently, a reference to the challenge of not excluding the merging of sustainable ways of life into the rural is recognised.

... order is only achieved through spatially open structures that are subject to change over time.

... future is designed as an open horizon, as a stylised space of possibilities that integrates diverse forms of life and economy into a German landscape that is considered as typical.

Conclusion

Through interpreting the case study, we were able to use the methodological approach (syntax-semantics-pragmatics) to show the extent to which new semantics of space and time can be retrieved through visualisations of generative AI. Although we were only able to illustrate the example of landscape here as an introduction, the extent to which context has already been added syntactically at the prompt level shows the ability of generative AI to generate and understand primitive semantics. Specifically, this opens up political spaces for the construction and control of extreme right-wing parties in order to virtually generate and disseminate distinctions as semantics that are stylised as worthy of preservation and protection. With regard to the visualisation results of the AfD party programme (2021), landscape serves as an anchor and traditional habitat for the community that is considered typical. Landscape will continue to be designed as an original construct in the sense of inverse landscapes; the change of vocabularies is therefore excluded. In contrast, the neutral prompt for generating the graphics in Figure 2 shows an open, harmonious development of landscape as the result of social and thus temporal change, which is spatially defined.

Of course, these graphics also necessarily involve the reduction of complexity and are equally contingent products of social self-observation and self-description. However, the difference between closeness (landscape as a fixed spatial construct that endures over time (Fig. 1)) and openness (landscape as a spatial construct that is subject to temporal change (Fig. 2)) becomes apparent. We are thus following Wardenga's observation that, particularly in the (right-wing) populist discourses of the present time, there is an increasing use of the concept of landscape - in other words, landscape ideals are linked to socio-structural imaginations (Korf et al., 2024, p. 11). This at least limits or even negates openness regarding social dynamics.

The space constructed by the AI used can be described as predominantly to exclusively rural. Urban elements are nowhere to be found in the visualisation of the spatial statements of the AfD's programmatic texts. In the prompt without such a reference, elements that can be described as urban can be found in the background or - in the 'vision of the future' -embedded in a rural context, particularly in the form of authorist architecture. Both tie in with a concept of romantic landscape that is widespread in the German-speaking world and is located beyond urban centres. The landscape view of urban and suburban spaces has only developed with deindustrialisation since the 1970s and is still practised in particular by people with a higher level of education (Hauser, 2001; Kiihne, 2018; Vicenzotti, 2011).

It would also be moreover interesting to create visualisations of landscape with reference to the party programmes of other centrist or left-wing parties and to contrast these with landscape-related visualisations based on the AfD party programme. This could also provide an insight into whether AI could also be used to generate visualisations that would be less characterised by the utopia/dystopia distinction, as the neutral prompt can already be read strongly as a Utopian design. In order to realise this, however, we would have to be able to draw on a comparable data set, such as the explorative study by Frankenberger et al. (2024). Such a comparative study and corresponding visualisations using the approach presented here (syntax-semantics-pragmatics) would also enable us to determine the extent to which LLMs are limited, the extent to which hallucinations of these systems can be provoked and the extent to which visual outputs are biased in such a way that they tend to generate extremes. This linking of syntax and semantics generally represents a current challenge in the research of AI and deserves special attention. By approaching this field, we could also investigate the extent to which algorithms develop the agency to produce actions outside of digital communication (pragmatics). To what extent, for example, would the reception of visualised idylls or nostalgia encourage various actors to take action and attempt to resist an open and Utopian vision of the future? In view of the political events of the present, a scientific examination of these topics is urgent. The triangulation of neopragmatist philosophy of language and system-theoretical social theory seems promising to us here and the methodological approach via syntax-semantics-pragamtics is equally expedient - from our perspective, this is a fruitful combination of abstract theorising and pragamtic research.

The results of our empirical study underline the urgency of developing a social understanding of these modes of irritation (see 1). This includes a pedagogical-didactic approach to this field - as a proven tool for addressing social challenges (Lammchen, 2024b, pp. 13-34) - in (political) education (in the public sector) as well as the necessary reflexive private handling of this access to new forms of social self-description. We must realise that we are acting as observers of our own observations and are called upon to reflect on the inherent (virtual) contingency in these processes. If we fail to do so, the consequences for a democratic, liberal society characterised by solidarity are hard to predict. Maintaining openness in the encounter with this media-generated complexity is only possible if we understand the degree of contingency in these construction processes. According to Rorty (2009 [1989], p. 48), this understanding as a constantly required cognitive achievement should be "the chief virtue of the members of a liberal society, and [...] the culture of such a society should aim at curing us of our 'deep metaphysical need'." Consequently, it is now easier than ever before to visualise these needs, which can be exposed as contingent, through generative AI or to retrieve them textually - however, we must always be aware of the contingency of this ourselves, i.e. we must reflexively check descriptions by observing observation and become our own sharpest critics. Only then can we deconstruct propagated truths, treat irritations as such, and maintain the openness that is essential to developing vocabularies and therefore society. An open society would then have the necessary space to endure time (Popper, 2011 [1945]).

References

References

Arnoldi, J. (2001). Niklas Luhmann. Theory, Culture & Society, 18(1), 1-13. https://doi.org/10.1177/02632760122051607

Baecker, D. (2018). 4.0 oder die Liicke die der Rechner Idsst. Merve-Verlag.

Bateson, G. (1985). Okologie des Geistes: Anthropologische, psychologische, biologische und epistemologische Perspektiven. Suhrkamp.

Beer, D. (2017). The social power of algorithms. Information, Communication & Society, 20(1), 1-13. https://doi.org/10.1080/1369118X.2016.1216147

Berr, K. (2019). Heimat und Landschaft im Streit der Weltanschauungen. In M. Hiilz, 0. Kuhne, & F. Weber (Eds.), Heimat: Ein vielfdltiges Konstrukt (pp. 27-51). Springer VS.

Berr, K., & Kuhne, 0. (2020). "Und das ungeheure Bild der Landschaft...": The Genesis of Landscape Understanding in the German-speaking Regions. Springer VS.

Berr, K., & Schenk, W. (2019). Begriffsgeschichte. In 0. Kuhne, F. Weber, K. Berr, & C. Jenal (Eds.), Handbuch Landschaft (pp. 23-38). Springer VS.

Bork-Hutter, I., Fuller, H., & Straube, 1. (Eds.). (2021).HanabuchDigitate Geographien: Welt - Wissen - Werkzeuge. Brill Schoningh.

Bourdieu, P. (1984). Distinction: A Social Critique of the Judgement of Taste. Harvard University Press.

Bruns, D., & Miinderlein, D. (2024). Interkulturelle Konstruktion. In 0. Kiihne, F. Weber, K. Berr, & C. Jenal (Eds.), Handbuch Landschaft (2nd ed., 609-616). Springer VS.

Burckhardt, L. (2006). Warum ist Landschaft schon? Die Spaziergangswissenschaft (M. Ritter, & M. Schmitz, Eds.). Martin Schmitz Verlag.

Calder, G. (2003). Rorty: AndRedesciption. Weidenfeld and Nicolson.

Denzer, V. (2024). Landschaft als Text - Textualitat von Landschaft. In 0. Kiihne, F. Weber, K. Berr, & C. Jenal (Eds.), Handbuch Landschaft (2nd ed., 143-150). Springer VS.

Doling, J., & Thielmann, T. (2009). Mediengeographie: Fur eine Geomedienwissenschaft. In J. Doling & T. Thielmann (Eds.), Mediengeographie: Theorie - Analyse - Diskussion (pp. 9-64). transcript.

Esposito, E. (1993). Der Computer als Medium und Maschine. Zeitschrift Fiir Soziologie, 22(5), 338-354-

Esposito, E. (2022). Artificial Communication: How Algorithms Produce Social Intelligence. The MIT Press.

Foerster, H. von (Ed.). (1995). Cybernetics Of Cybernetics: The Control Of Control And The Communication Of Communication (2nd ed.). Future Systems.

Foucault, M. (1996). Diskurs und Wahrheit: Berkeley-Vorlesungen 1983. Merve-Verlag.

Foucault, M. (2007 [frz. Original 1971]. Die Ordnung des Diskurses: Mit einem Essay von Ralf Konersmann. Fischer.

Franke, N. M. (2019). Umweltschutz ist Heimatschutz? Der Zugang des rechtsextremistischen Denkens zum Thema Heimat und eine demokratische Gegenposition. In M. Hiilz, 0. Kiihne, & F. Weber (Eds.), Heimat: Ein vielfdltiges Konstrukt (pp. 391-401). Springer VS.

Frankenberger, R., Hinz, L., Kiihne, 0., Pfau, B., & Schmid, E. (2024). Raumkonstruktionen extrem rechter Parteien in Deutschland: Eine explorative Studie. Springer VS.

Glasersteld, E. v. (1988). Eintuhrung in den radikalen Konstruktivismus. In P. Watzlawick (Ed.), Die erfundene Wirklichkeit: Wie wissen wir, was wir zu wissen glauben?Beitrage zum Konstruktivismus (5th ed., pp. 16-38). Piper.

Glasersfeld, E. v. (1995). Radical constructivism: A way of knowing and learning. Studies in mathematics education series: Vol. 6. Falmer Press.

Glasersfeld, E. v. (2010). Konstruktion der Wirklichkeit und des Begriffs der Objektivitat. In H. Gumin & H. Meier (Eds.), Verbffentlichungen der Carl-Friedrich-von-Siemens-Stiftung: Vol. 5. Einfuhrung in den Konstruktivismus: Beitrage von Heinz von Foerster, Ernst von Glasersfeld, Peter M. Hejl, Siegfried J. Schmidt, Paul Watzlawick (12th ed., pp. 9-40). Piper.

Goeke, P., & Lippuner, R. (2011). Geographien sozialer Systeme. Soziale Systeme -Zeitschrift Fiir Soziologische Theorie, 17(2), 227-233. https://doi.org/10.5167/uzh-73942

Goeke, P., Lippuner, R., & Wirths, J. (2015). Von der Geographie sozialer Systeme zu einer allgemeinen Okologie der Gesellschaft: Eine Einleitung. In P. Goeke, R. Lippuner, & J. Wirths (Eds.), Konstruktion und Kontrolle: Zur Raumordnung sozialer Systeme (pp. 9-22). Springer VS.

Hard, G. (1969). Das Wort Landschaft und sein semantischer Hof: Zu Methode und Ergebnis eines linguistischen Tests. Wirkendes Wort, 19, 3-14.

Hauser, S. (2001). Metamorphosen des Abfalls: Konzepte fiir alte Industrieareale. Campus.

Hokema, D. (2013). Landschaft im Wandel? Zeitgenossische Landschaftsbegrijfe in Wissenschaft, Planung undAlltag. Springer VS.

Horster, D. (1991). Richard Rorty zur Einfiihrung (1. Aufl.). Zur Einfiihrung: Vol. 68. Junius.

Ipsen, D. (1992). Stadt und Land - Metamorphosen einer Beziehung. In H. HauBermann, D. Ipsen, R. Kramer-Badoni, D. Lapple, M. Rodenstein, & W. Siebel (Eds.), Stadt und Raum: Soziologische Analysen (2nd ed., pp. 117-156). Centaurus.

Jaworski, A., & Thurlow, C. (2010). Introducing Semiotic Landscape. In A. Jaworski & C. Thurlow (Eds.), Semiotic Landscapes: Language, Image, Space (pp. 1-40). Continuum International Publishing Group.

Kirchhoff, T., & Trepl, L. (2009). Landschaft, Wildnis, Okosystem: zur kulturbedingten Vieldeutigkeit asthetischer, moralischer und theoretischer Naturauffassungen: Einleitender Uberblick. In T. Kirchhoff & L. Trepl (Eds.), Vieldeutige Natur: Landschaft, Wildnis und Okosystem als kulturgeschichtliche Phanomene (pp. 13-68). transcript.

Korf, B., RothfuB, E., & Wardenga, U. (2024). Disziplinhistorische Tauchgange zur German Theory: Ein Gesprach mit Ute Wardenga iiber die deutsche Landerkunde und Landschaftsgeographie. Geographica Helvetica, 79(S), 1-13. https://doi.org/10.5194/gh-79-1-2024

Korner, S. (2006). Heimatschutz, Naturschutz und Landschaftsplanung. In Institut fur Landschaftsarchitektur und Umweltplanung - Technische Universitat Berlin (Ed.), Perspektive Landschaft (pp. 131-142). Wissenschaftlicher Verlag.

Krotz, F. (2012). Mediatisierung als Metaprozess. In J. Hagenah & H. Meulemann (Eds.), Mediatisierung der Gesellschaft? (pp. 19-41). LIT.

Krotz, F. (2015). Mediatisierung: Fallstudien zum Wandel von Kommunikation. In A. Hepp, F. Krotz, S. Lingenberg, & J. Wimmer (Eds.), Handbuch Cultural Studies und Medienanalyse (pp. 439-451). Springer VS. https://doi.org/10.1007/978-3-531-90414-6

Kiihne, 0. (2013a). Landschaft zwischen Objekthaftigkeit und Konstruktion -Uberlegungen zur inversen Landschaft. In D. Bruns & 0. Kiihne (Eds.), Landschaften: Theorie, Praxis und internationale Beziige: Impulse zum Landschaftsbegrijfmit seinen dsthetischen, okonomischen, sozialen und philosophischen Beziigen mit dem Ziel, die Verbindung von Theorie undPlanungspraxis zu starken (pp. 181-193). Oceano Verlag.

Kiihne, 0. (2013b). Landschaftsasthetik und regenerative Energien - Grunduberlegungen zu De- und Re-Sensualisierungen und inversen Landschaften. In L. Gailing & M. Leibenath (Eds.), Neue Energielandschaften - Neue Perspektiven der Landschaftsforschung (pp. 101-120). Springer VS.

Kiihne, 0. (2018). Landscape and Power in Geographical Space as a Social-Aesthetic Construct. Springer International Publishing.

Kiihne, 0. (2019). Landscape Theories: A Brief Introduction. Springer VS.

Kiihne, 0. (2021). Potentials of the Three Spaces Theory for Understandings of Cartography, Virtual Realities, and Augmented Spaces. KN - Journal of Cartography and Geographic Information, 71(4), 297-305. https://doi.org/10.1007/s42489-021-00089-w

Kiihne, 0. (2024a). Autopoietische Systemtheorie und Landschaft. In 0. Kiihne, F. Weber, K. Berr, & C. Jenal (Eds.), Handbuch Landschaft (2nd ed., 151-163). Springer VS.

Kiihne, 0. (2024b). Redescribing Horizontal Geographies: A Neopragmatist Approach to Spatial Contingency, Complexity, and Relationships. Springer International.

Kiihne, 0., & Berr, K. (2024). Romantikund Landschaft. In 0. Kiihne, F. Weber, K. Berr, & C. Jenal (Eds.), Handbuch Landschaft (2nd ed., 77-93). Springer VS.

Kiihne, 0., Berr, K., Edler, D., Lohmann, P., & Schuster, K. (2025). Neopragmatismus -inverse Landschaft - (karto)graphische Darstellung: Von der Kritik zur Weiterentwicklung. Springer VS.

Kiihne, 0., Berr, K., & Koegst, L. (2023). Contingency and Landscape: Basic Considerations on Graphic and Cartographic Representations in Recourse to the Concept of Inverse Landscapes as a Contribution to Deviant Cartographies with Examples on Louisiana. KN - Journal of Cartography and Geographic Information, 1-12. https://doi.org/10.1007/s42489-023-00145-7

Kiihne, 0., & Edler, D. (2022). Georg Simmel Goes Virtual: From 'Philosophy of Landscape' to the Possibilities of Virtual Reality in Landscape Research. Societies, 12(5), 122. https://doi.org/10.3390/soc12050122

Kiihne, 0., Edler, D., & Jenal, C. (2021). A Multi-Perspective View on Immersive Virtual Environments (IVEs). ISPRS - International Journal of Geo-Information, 10(8), Article 518,1-22. https://doi.org/10.3390/ijgi10080518

Laclau, E., & Mouffe, C. (1985). Hegemony and Socialist Strategy: Towards a Radical Democratic Politics. Verso.

Lammchen, R. (2023). Observing Landscape: A Systems Theoretical Approach. In L. Koegst, 0. Kiihne, & D. Edler (Eds.), Multisensory Landscapes: Theories, Research fields, Methods - an Introduction (73-89). Springer Fachmedien.

Lammchen, R. (2024a). Algorithmen als Interaktionspartnerund Beobachter: Die Perspektivierung von Natur in Social Media. In S. Feige & N. S. Richter (Eds.), Sozial-und Kulturgeographie: Vol. 63. Natur und Bildung: Interdisziplinarefachdidaktische undpadagogische Perspektiven (pp. 29-52). transcript.

Lammchen, R. (2024b). Vom Bud zur Bildung: Computerspiele als raumkonstituierende Bildungsmedien der Gegenwart. Springer, https://doi.org/10.1007/978-3-658-46575-9

Lammchen, R. (2025). Algorithmic Art: Neo-deviant presentations of observations of the present exemplified by the creation of generative landscape art with Midjourney. In F. Papadimitriou & 0. Kiihne (Eds.), Deviant Landscapes: A Journey to Exotic and Imaginary Places and Spaces (pp. 441-456). Springer VS.

Lippuner, R. (2007). Kopplung, Steuerung, Differenzierung: Zur Geographie sozialer Systeme. Erdkunde, 61(2), 174-185.

Lohmann, P., Kiihne, 0., & Berr, K. (2024). Landschaft und Gefiihl - eine neopragmatistische Redeskription auf Grundlage der Philosophie Fichtes: Ein Beitrag zur humanistischen Geographie. SpringerVS.

Luhmann, N. (1984). Soziale Systeme: Grundrifi einer allgemeinen Theorie. Suhrkamp.

Luhmann, N. (1989 [1980]. Gesellschaftsstruktur und Semantik: Studien zur Wissenssoziologie der modernen Gesellschaft (Vol. 1). Suhrkamp.

Luhmann, N. (1990). Die Wissenschaft der Gesellschaft. Suhrkamp.

Luhmann, N. (1993 [1980]. Gesellschaftsstruktur und Semantik: Studien zur

Wissenssoziologie der modernen Gesellschaft (Vol. 1). Suhrkamp. Luhmann, N. (1997). Die Gesellschaft der Gesellschaft. Suhrkamp.

Luhmann, N. (2016 [1999]. Gesellschaftsstruktur und Semantik: Studien zur Wissenssoziologie der modernen Gesellschaft (3rd ed., Vol. 4). Suhrkamp.

Luhmann, N. (2018 [1998]. Die Gesellschaft der Gesellschaft (10th ed.). Suhrkamp.

Lundby, K. (2014). Mediatization of Communication. In K. Lundby (Ed.), Mediatization of Communication (pp. 3-35). de Gruyter.

Maturana, H. R., & Varela, F. J. (1987). The Tree of Knowledge: The Biological Roots of Human Understanding (Revised edition). Shambhala.

Micheel, M. (2012). Alltagsweltliche Konstruktionen von Kulturlandschaft. Raumforschung Und Raumordnung - Spatial Research and Planning, 70(2), 107-117. https://doi.org/10.1007/s13147-011-0143-x

Morris, C. W. (1938). Foundations of the Theory of Signs. Universitat of Chicago Press.

Miiller, G. (1977). Zur Geschichte des Wortes Landschaft. In A. Hartlieb von Wallthor & H. Quirin (Eds.), "Landschaft" als interdisziplindres Forschungsproblem: Vortrage und Diskussionen des Kolloquiums am 7-/8. November 1975 in Miinster (pp. 3-13). Aschendorff.

Miiller, M. (2021). Rorty lesen. Springer VS.

Miiller, M. (Ed.). (2023). Handbuch Richard Rorty. Springer Fachmedien. https://doi.org/10.1007/978-3-658-16253-5

Miiller-Salo, J. (2020). Klima, Sprache und Moral: Eine philosophische Kritik. Philipp Reclam jun.

Nassehi, A. (2008). Die Zeit der Gesellschaft: Auf dem Weg zu einer soziologischen Theorie der Zeit. Neuauflage mit einem Beitrag "Gegenwarten" (2nd ed.). Springer VS. https://doi.org/10.1007/978-3-531-91099-4

Nassehi, A. (2019). Muster: Theorie der digitalen Gesellschaft (3rd ed.). C.H. Beck.

Nassehi, A. (2021). Muster: Theorie der digitalen Gesellschaft. Beck.

O'Grady, W., & Lee, M. (2023). Natural Syntax, Artificial Intelligence and Language

Acquisition. Information, 14(7), Article 418. https://d01.org/10.3390/1nf014070418 Peirce, C. S. (1991). Schriften zum Pragmatismus und Pragmatizismus (K.-O. Apel, Ed.). Suhrkamp.

Peirce, C. S. (1998). Essential Peirce: Selected Philosophical Writings. VOLUME 2 (1893-1913) (the Peirce Edition Project, Ed.). Indiana University Press.

Piechocki, R. (2010). Landschaft - Heimat - Wildnis: Schutz der Natur - aber welcher und warum? Beck.

Popper, K. R. (2011 [1945]. The Open Society and Its Enemies. Routledge.

Putnam, H. (1990). Die Bedeutung von "Bedeutung" (2., durchgesehene Auflage). Klostermann.

Redepenning, M. (2006). Worn Raum? Systemtheorie, critical geopolitics und raumbezogene Semantiken. Beitrage zur Regionalen Geographic: Vol. 62. Selbstverlag.

Redepenning, M., & Wilhelm, J. (2014). Raumforschung mit luhmannscher Systemtheorie. In J. OBenbriigge & A. Vogelpohl (Eds.), Theorien in der Raum- und Stadtforschung: Einfuhrungen (pp. 310-327). Westfalisches Dampfboot.

Reese-Schafer, W. (2016). Richard Rorty zur Einfuhrung. Junius.

Rorty, R. (1997 [1989]. Contingency, Irony, and Solidarity. Cambridge University Press. https://doi.org/10.1017/CBO9780511804397

Rorty, R. (1998). The Contingency of Language. In M. F. Bernard-Donals & R. R. Glejzer (Eds.), Rhetoric in an antifoundational World: Language, culture, and pedagogy (pp. 65-85). Yale University Press.

Rorty, R. (2003). Wahrheit undFortschritt. Suhrkamp.

Rorty, R. (2009 [1989]. Contingency, irony, and solidarity (28th ed.). Cambridge University Press.

Rorty, R. (2023). Pragmatismus als Antiautoritarismus (J. Schulte, Trans.) (E. Mendieta, Ed.). Suhrkamp.

Schenk, W. (2013). Landschaft als zweitache sekundare Bildung - histonsche Aspekte 1m aktuellen Gebrauch von Landschaft im deutschsprachigen Raum, namentlich in der Geographie. In D. Bruns & 0. Kuhne (Eds.), Landschaften: Theorie, Praxis und internationale Beziige: Impulse zum Landschaftsbegrijf mit seinen dsthetischen,

okonomischen, sozialen und philosophischen Beziigen mit dem Ziel, die Verbindung von Theorie und Planungspraxis zu stdrken (pp. 23-36). Oceano Verlag.

Schenk, W., & Steinkriiger, J.-E. (2020). Kulturlandschaft als kultureller Prozess, illustriert am Orden der Zisterzienser und an den friihen Tragern der Rheinromantik. In R. Duttmann, 0. Kiihne, & F. Weber (Eds.), Landschaft als Prozess (pp. 119-134). Springer VS.

Schlottmann, A., & Miggelbrink, J. (2015). Ausgangspunkte: Das Visuelle in der Geographie und ihrer Vermittlung. In A. Schlottmann & J. Miggelbrink (Eds.), Visuelle Geographien: Zur Produktion, Aneignung und Vermittlung von RaumBildern (pp. 13-25). transcript.

Sieferle, R. P. (1985). Heimatschutz und das Ende der romantischen Utopie. Arch+, 81, 38-42.

Spencer Brown, G. (1969). Laws of Form. Allen & Unwin.

Topper, K. (1995). Richard Rorty, Liberalism and the Politics 01 Redescnption. American Political Science Review, 89(4), 954-965. https://doi.org/10.2307/2082520

Varela, F. J., Maturana, H. R., & Uribe, R. (1974). Autopoiesis: The organization of living systems, its characterization and a model. Biosystems, 5(4), 187-196. https://doi.org/io.ioi6/0303-2647(74)9003i-8

Vicenzotti, V. (2011). Der "Zwischenstadt"-Diskurs: Eine Analyse zwischen Wildnis, Kulturlandschaft und Stadt. transcript.