A Methodology for the Study of Interspecific Cohabitation Issues in the City

The present article will introduce a proposition of semiotic methodology that can be used to diagnose cohabitation issues in cities between human inhabitants and non-human liminals. This methodology is built on a few sets of data that should be easy to obtain in any important city, and can therefore be utilised in a variety of situations. The different sets of data allow us to map the cohabitation semiosphere (following Hoffmeyer’s meaning of the term) of the situation along three axes: the materiality of the situation, the symbolic significance of the relationship, and the emotional significance of the interaction. These three aspects allow us to see gaps, paradoxes and points of consistency, enabling complex and multi-level understanding of the situation.


Theoretical Frame
Historically speaking, semiotics was concerned with living beings and their natural behaviours and interactions with their environment. Semiotics as a tool to study living beings re-emerged during the twentieth century with Sebeok's work on zoosemiotics (Sebeok, 1968(Sebeok, , 1972(Sebeok, , 1993, as a way to understand animal meaning-making, using Peirce's sign theory (1868( , Sebeok, 1994.
Aside from this, studying animals with linguistic tools was not common in the academic fields, especially among humanities, and a certain hostility toward such studies remained for a long time (Griffin, 1977;De Waal, 2014). Yet, ethology had for a long time craved tools to research inside the "black box" of the mind (Eco, 2000), to understand how animals interacted with their environment, how they understood it and lived in it, and how to understand what Uexküll named the Umwelt (Uexküll, 2010, Kull, 1998.
Biosemiotics and zoosemiotics will be used in this project in a relevant but quite rare way, as far as empirical work is concerned: as tools and models allowing a study in both human and animal Umwelten, especially regarding how humans perceive animals -corvids in this study -how corvids behave in a human environment -in this paper the city of Tartu -and how they interact with each other. There are already many theoretical works on this aspect that use different main concepts to propose a lecture grid of human-animal interactions: eco-field (Farina & Belgrano, 2006), semiosphere (Maran, 2021), critical zoosemiotics (Martinelli, 2019). But, with some notable exceptions (Jaroš, 2018(Jaroš, , 2021, few works in semiotics are firstly and mainly empirical. The approach introduced was chosen to create the most practical methodology possible, using available data sets and building a semiotic model that will use what is possible to gather from very basic and simple empirical projects. There currently are very few studies in comparative ethology or semiotics of culture applied to animals in Europe in an empirical and practical aspect, even if this is evolving (Waal, 2017). While the method has begun to be applied in primatology in Japan (Matsuzawa, 2017), few species have been the subject of such studies apart from great apes, and even fewer are included in projects aiming to understand a whole ecosystem with its interactions and its tensions, even of this kind of work does exist in a more theoretical aspect (Tønnesen & Rattasepp, 2016).
The urban environment, while well-known, is still barely studied from this perspective. Nonetheless, it has a rich biodiversity, and with the development of urbanisation and increasing population density, urban environments are a critical aspect of the coexistence between humans and other species. Consequently, it makes it necessary to better understand this environment and how to manage it with a more global vision. To do so, semiotics models and tools will be used with other academic specialities -mostly ethology, regarding data and collection methods 1 -and theories -especially relevance theory (Sperber & Wilson, 2001) according to the method of intertheoricity (Guillaume 2014b), which allows disciplines of different fields to work correctly together on complex problems, like animal tastes and preferences (Kreutzer & Aebischer, 2015), emotions (Delahaye, 2019) or sensitivity (Guillaume, 2014a).

Research Aims and Question
This paper aims to introduce a semiotic methodology that can diagnose interspecies cohabitation issues, by mapping the local semiosphere (following Hoffmeyer's mean-1 In the case of the Tartu empirical study, two major ethological works were used: the noise repellents experimentation requested by the Tartu city government (report available in references), and the data from the bird team of the Bioveins Project (these ones were kindly shared by the team before the publication of their paper and cannot be communicated, but contact is in references for any needed request).
ing 2 ). It will show how easy it can be to obtain data that can be used in a semiotic frame to draw a relevant and useful summary of a cohabitation situation, its different aspects and, depending on the situation, its issues and possible ways to solve them.
This methodology is the result of two different empirical studies about cohabitation between humans and liminal animals in the urban context. The first one was set up in Paris between 2019 and 2020, and was focused on rats and mice (Delahaye, 2021). Due to the beginning of the Covid pandemic and several lockdowns in Paris, it has some lacking. The second one started in 2021 in Tartu, focusing this time on corvids, and is still ongoing (Delahaye, 2023). This paper will try to present methodology built from these previous studies in the most generalist way possible, with the aim of generalizing to other fields, with other species, for other research teams.

Methods and Materials
Biodiversity is a major issue in our society. Yet, even if an important part of the population is aware of that issue, the cities are usually not seen as an ecosystem and the species living with them are not considered as part of the local biodiversity. The gap between this perception and reality has an impact on how we think about and manage animals (Forsberg et al., 2012).
The current project focused on corvids (Corvus cornix, Corvus frugilegus, Coloeus monedula) in Tartu and partially followed the methodology used in a previous study (Delahaye, 2021) of another urban species (Rattus norvegicus) in the urban environment of Paris, France, to propose a way to create a mapping of the situation and a "diagnosis" of potential issues. But the proposed methodology should be relevant to address any non-threatening cohabitation situation (this methodology was not built to address situations where human lives will be immediately at risk).

Methodology
The mapping of a situation of cohabitation requires three poles: the materiality of the situation, the symbolic significance of the situation, and the emotional significance of the situation.
Materiality this pole aims to describe factual elements of the current situation. How is the species fairing in the city? Is it in good health, is there an overpopulation issue, what presence (i.e., begging for food) or absence (i.e., annual migration) is recorded? This part is important to create a comparison point with other aspects. For example, if three cases of aggressive behaviour are recorded, they will not have the same importance if the global population is estimated at around a few dozen or a few thousand individuals, and the focus inhabitants will have on these cases will be more or less justified. This aspect is obtained by comparing biodiversity data from two different sources: an academic data set (often fragmented but with solid methodology, which has strong reliability, but is less encompassing) and a data set from citizen science (often very numerous but with fragile methodology, which has strong power, but less strong reliability).
The Symbolic Significance of the Situation this pole aims to describe the symbolic aspect of the studied species. Inhabitants can very well be unaware of the symbolic features they attribute to the species. It is, consequently, important to describe them, as they can be the origin of behaviours or mindsets that are, without this aspect, difficult to understand. For example, if inhabitants are perfectly aware of the nonhazardous quality of a species, but still declare a recalcitrant fear of it, the origin can be found in this pole. This aspect is obtained by comparing the academic data set about the behaviour and biology of the species (data that are the most neutral and reliable) to the textual data set (data about how people talk about the species and about the kind of descriptions of the species they can be exposed to in their cultural environment). Very strong textual patterns or narrative items 3 which are not congruent or even contradictory with ethological data are the marks of purely symbolical significations, existing next to the factual data, often with very few interactions.
The Emotional Significance of the Situation this pole aims to describe the emotional relationship inhabitants have with the studied species. Inhabitants are usually very aware of this aspect, and it can be in relation to a paradox with a symbolic aspect. For example, inhabitants can have, in the same movement, fear or disgust for a species, due to its symbolic significance, but not want any harm done to them (and they may even strongly oppose extermination campaigns) due to their emotional significance. This aspect is obtained by comparing the data set from citizen science (where the difference between inhabitant measurements and academic ones can show the biases they have toward this species) to the textual data again (to gather how people talk about or what they are used to reading or hearing about the species). These two aspects will usually allow the creation of an accurate representation of the inhabitants' feelings toward the species.
With these three poles, a mapping of the total semiotic situation can be done. Making these three poles interact will help to pinpoint where exactly an issue is. Is it a material issue, that must be addressed through practical methods (in urban planning, city management etc.)? Is it an emotional issue, that can be addressed through awareness campaigns or popularisation events, to make people feel safe or to develop their empathy? Is it a symbolic issue, that should be addressed by proposing cultural content and new narratives to feed people's symbolic imagination or references? Of course, these aspects can never be fully and artificially separated since, as shown in Fig. 1, they interact closely with each other. But understanding the main issue, what aspect is the root of the others or which one is currently the most urgent and sensitive is the best way to build relevant and accurate solutions.

Materials
The three poles required to map the situation of cohabitation emerge from data sets. These sets of data are conceived in such a way that they can be possible to create in almost every big city or urban centre. There are three families of sets: sets on academic research on biodiversity, sets from citizen science programs, and sets from textual materials. The comparison between the two data sets will lead to the emergence of one of the three aspects of the local semiosphere: materiality, emotional significance and symbolic significance. No set of data should be seen as the chief source of an aspect, except in situations where an important difference of quality (in number, reliability or novelty of the data) exists between two sets and this difference is not possible to be solved at the time of the project. There, the most reliable set should be seen as the chief source.
The family of biodiversity data must include at least two studies or groups of studies. They should be from different teams or projects and should have different scopes or purposes. In the Tartu project, the first study was mainly focused on nuisances and repellents, whereas the second was a comparative census of biodiversity, studying different cities all around Europe. Having different teams and scopes or purposes is a good way to be sure that the consistency of data is not an artefact. The main limitation of this kind of data is that academic research is usually limited: a study can be done only for a couple of years, projects usually involve a small number of researchers, who can only be at one or two places at a time. This is the main reason why, to grasp the materiality of the situation, these data are not sufficient on their own. They should Fig. 1 Different aspects of the semiosphere in a situation of interspecific cohabitation emerging from data sets be compared with the ones emerging from citizen science sets of data, which are less reliable but more encompassing.
The family of citizen science data must include at least two different programs, sources or databases. Ideally, one should be as free and open as possible, and one should be more strictly framed by professionals. This variation should allow the checking of data consistency between these sets and the academic ones, but should also allow spotting gaps between what should be recorded by citizens, and what they feel should be recorded. The more open and freer the program or database is, the more important it is to be very cautious about its data and to consider it more like a record of how inhabitants perceive animals and biodiversity around us, rather than an accurate record of animals and biodiversity in this area. The gaps appearing in these sets of data, when compared to textual data, can enlighten us on the emotional significance of the studied species for inhabitants.
The family of textual data can be particularly difficult to compound, as it should include mainly texts in the native language of the studied inhabitants. It should include both day-to-day texts (blogs, complaints, interviews with inhabitants etc.) and more literary material. The aim here is to draw a comprehensive summary of the collective symbolical aspect of the animal, its traditional meaning, and its role in folklore or the collective unconscious. The material should consequently be adapted and gathered where it is relevant. If the animal is linked to an important literary, folkloric or aesthetic background, texts from academics providing analysis of this background should also be included, especially if the literary texts are difficult to access for linguistic reasons. The comparison of these sets with the ones of other families is a good way to spot paradoxical situations. For example, if the narrative emerging from the data does not match the scientific data about behaviour and biodiversity, it is a sign that the symbolical significance of the species is stronger than the facts people can be exposed to. This kind of situation will require very different solutions than situations where the two families of data are consistent, and the issue is more on the emotional aspect.

Results
Comparison between the materiality of the situation and emotional or symbolic significance will show how the perceived or "ideal" animal differs from the biological animal in a context of cohabitation. This step is critically important to understand what should be addressed to solve a potential issue, as the issue may not be what we observe in the fieldwork but what inhabitants perceive as an issue.
For example, in the situation of Tartu, it appears that the gap between an "ideal corvid" and the biological reality of the species could be part of the explanation of why corvids' observations are so sensitive to remarkable bias. The symbolical significance given to the species impacts the way inhabitants perceive, register, and react to the materiality of the species. Three elements can be pointed out: • Corvids are perceived as a prolific and almost overpopulating species: therefore, they are poorly registered through citizen science programs because the general public sees no point to do so. • Corvids are perceived as a source of nuisances, even if experimentations done on this aspect show that these nuisances are minimal (they are less noisy than repellents used against them), and even if these nuisances are in fact a completely natural part of their social behaviour (no "pathological" behaviour is really registered, like attacking humans). • The homogenous presence of the species in the city leads to inhabitants considering them as "objects parts of the city" (this shows especially when studying how the birds are described as ingredients of a city's atmosphere, closely linked to the buildings they live around), rather than as actual animals and part of the city's biodiversity (which probably explains their underrepresentation in the biodiversity's watch database 4 ). This bias is well-known by scientists who insist strongly, in citizen science programs, on the necessity to record all species, including the most common ones 5 .
If a gap between the materiality of the situation and its emotional and/or symbolic significance appears, another step should take place. A comparison between elements of symbolic significance and emotional significance will show where exactly a paradox is occurring and will give different ways to access the inhabitants' semiotic representations to solve the issue. Usually, symbolic significance is transmitted by culture, tradition, history, folklore, or even more unconscious ways, and it can be hard to address. But emotional significance is usually more accessible, especially through new narratives aiming at raising awareness or developing empathy. For example, in Tartu, the quite paradoxical elements listed just before are probably enforced by another kind of gap, rooted in the emotional significance corvids seem to have, at least for part of the inhabitants. The importance of the remarkable bias, and, consequently, the negligible place that the inhabitants seem to grant to corvids, is apparently contradictory to the importance given to the nuisances. This apparent paradox can be explained by two other mapping elements: • The underrepresentation of biological individuals in the citizen science database and overrepresentation of complaints about normal biological behaviours, which nuisances are objectively very mild, are in fact coherent when considering that corvids are seen by a part of inhabitants as "objects of the city": as objects, they are not registered like other living species, but they are also not supposed to create nuisances for inhabitants. • The underrepresentation of biological individuals in the citizen science database is also relevant for the group of inhabitants that give a positive emotional significance to corvids, as this positive emotional significance is also closely related to the concept of "objects of the city". This positive emotional significance is visible in the commitment inhabitants have to the particular aesthetic of their city, of which corvids are an element.
During the study on rats in Paris, the materiality of the species was also very different from the other aspects, and a gap was present between the symbolical significance of the species and the emotional significance, with a notable amount of participants in the study agreeing on rats being nuisances but having empathy for them and not willing them any harm. In the current project, differences observed are more likely the sign of two different strong categories of inhabitants, having different kinds of emotional and symbolical perceptions of the corvids, but cohabitating in the same city: • The objectification profile: inhabitants with this profile perceive corvids as "object parts of the city" mostly due to their omnipresence. As perceived urban objects, corvids are not supposed to bother the inhabitants, which are consequently very sensitive to nuisances, or what is perceived as nuisances, from what they fail to perceive as living beings with biological behaviours and needs. These people can, on another hand, be bird enthusiasts regarding other species. If so, they can feed the citizen science database while being subject to the remarkable bias. • The sublimation profile: inhabitants with this profile perceive corvids as "objects parts of the city" mostly due to their contribution to the gothic aesthetic of Tartu. As atmosphere creators, corvids are perceived with a positive emotional significance, which includes aspects that can be seen as nuisances by the other profile, especially regarding the corvids' omnipresence and noises. If these people can also be bird watchers, they are still probably subject to the remarkable bias, as they still perceive corvids as urban objects.
Of course, other profiles of inhabitants are possible, but these two are probably the ones with the most important influence on the data. It is also possible that different profiles, or elements of the profile, coexist in the same person, due to contradictory experiences and the fact, already observed in Paris, that symbolic and emotional significances can be in a paradoxical situation inside the same person.

Discussion
The methodology addresses an increasingly frequent and critical problem with the progress of urbanization. The dangers which weigh on biodiversity are indeed numerous and, as the urban proportion of fauna increases, it becomes urgent to also take an interest in the protection and conservation of this specific part of biodiversity. Not all urban species are endangered, of course, but all are part of this ecosystem, and solving cohabitation issues is needed as a global measure in favour of biodiversity. This type of study is all the more important since cities are a privileged environment for these species, especially corvids (Marzluff et al., 2001, Chap. 16), and the main problems they encounter come from their cohabitation with humans.
The methodology is immediately interesting for possible exploitation. Theoretical and practical toolkits were currently tested, based on the results and their methodology. These toolkits were guidelines to improve the quality of citizen science programs. They are a toolkit for officials and local governments to choose non-lethal solutions in cohabitation issues and guidelines for communication teams to help set up the most efficient media depending on the identified issues. They aim to help, improve and facilitate the mapping of the semiosphere between inhabitants and liminals in a large variety of contexts, to easily and quickly find the gaps and the paradoxes that need addressing for cohabitation improvement. These toolkits take into account the work that other teams shared with us in their own attempt to solve local cohabitation issues by providing non-lethal methods to reduce nuisances caused by some species (Lequitte-Charransol & Jiguet, 2021) and improve the quality of the cohabitation.
Further research is currently ongoing to assess and demonstrate the ability of this methodology to be useful to various cities, cultures, and species. The results presented in this paper could be improved, modified or nuanced by future results obtained by another research team using this methodology.