This thesis is based on a case study of the degraded Boksburg Lake social-ecological system and an environmental education initiative that aimed to support its transformation. This initiative aimed to involve local people in reclaiming the lake’s social and ecological value, through a process of collectively reimagining possibilities, shaping identities, gaining knowledge and developing local human agency. The focus was on social learning processes in schools and churches to explore opportunities for co-engaged reflexivity that might produce transformation. Schools and Christian churches, two institutions that reflect modern, western socialecological worldviews also have the potential to bring about change. Critical Realism was chosen as my philosophical framework as it provided the tools to explore deeper mechanisms beyond empirical reality, both influencing the degrading trajectory as well as providing possibilities for transformation. It also legitimised case study research as a means to understand more generalised processes characterising modern social-ecological systems. The choice of Critical Realism informed the scope of my primary research question: What generative mechanisms constrain and enable the development of social-ecological resilience for well-being, in the modern social-ecological system of Boksburg Lake? The following three goals were formulated to address this primary question. Goal 1: Based on a multitheoretical perspective of social-ecological literature, develop conceptual tools that have explanatory power to probe generative mechanisms operating in the Boksburg Lake social-ecological system. Goal 2: Identify generative mechanisms driving the current degradation of the Boksburg Lake social-ecological system. Goal 3: Identify learning mechanisms that support transformation for greater social-ecological resilience of the Boksburg Lake social-ecological system. By addressing the primary question and research goals I aimed to gain insights into modern global socialecological systems, the mechanisms that drive high social-ecological risk and the requirements for and possibilities of global systemic change. Drawing on a broad reading of social-ecological literature from different vantage points, tools with explanatory power were developed to probe for generative mechanisms in the Boksburg Lake social-ecological system (goal 1). The human capacity for symbolic representation is identified as an emergent property of coevolving human-ecological systems. These symbolic representations become expressed in culture and worldviews, and influence patterns of identifying, types of knowledge and forms of agency. The nature of these will determine the degree that cultural systems are embedded within ecological reality and the extent of cultural-ecological coupling. A cultural system closely coupled with ecological realities is likely to value ecological systems and manage them for their health, while less coupled cultural-ecological systems are likely to lead to the opposite. Because of their integrated nature, the extent of ecological health and value will affect the decline or sustainability of cultural-ecological systems. There are numerous examples of the learning that can take place when cultural-ecological systems are facing decline. This learning can enhance or reduce biophyllic instincts that become encoded in patterns of identifying, types of knowledge and forms of agency. This in turn affects the strength of cultural-ecological coupling and the extent that human societies co-evolve with ecological systems. / Normalising ideologies is a concept coined in the thesis to refer to symbolic representations of reality that have become integral to a social fabric and determine meaning, while maintaining the domination of the powerful. These ideologies determine patterns of identifying, knowledge and agency and are recognised as having a fundamental influence on the resilience of social-ecological systems. Four normalising ideologies are identified that promote apparent human progress at the expense of ecological integrity and social equality and thus alienation with each other and the ecological world. These are human-ecological dualism, anthropocentrism, nature is mechanised and nature is to be controlled. There are also a number of ideologies promoting connectedness with the ecological world that, if they became normalised, would support greater social-ecological resilience for well-being. Generative mechanisms driving the current degradation of the Boksburg Lake socialecological system were identified (goal 2). Drawing on critical methodology, the main method adopted was document analysis of the Boksburg Advertiser archives, Boksburg’s local newspaper. Four generative mechanisms are recognised as most influential. Two of these have been named hegemonic symbolic systems. The primary symbolic system consists of the four normalising ideologies, mentioned above, that promote human progress at the expense of ecological health. The secondary, more explicit symbolic system, built on this, consists of the following normative ideologies: economic growth is imperative, unrestrained development is promoted, competition is the necessary means and consumerism is the good life. These two symbolic systems have had causal influence on the systematic erosion of ecological processes and biological diversity that has occurred in Boksburg, with the consequent undermining of social-ecological resilience for well-being. The third mechanism that constrains resilience is the power dynamics that have shaped Boksburg’s economic history and social-ecological system. This has resulted in a society built on inequality and injustice with all its associated social and environmental ills, expressed as externalities. The fourth mechanism resides in Boksburg’s political and municipal dynamics. These structures are not designed to tackle complex social-ecological problems and they hold considerable agential power, yet seem dysfunctional at present. Learning mechanisms that support transformation for greater social-ecological resilience of the Boksburg Lake social-ecological system were identified (goal 3). By adopting the role of a reflexive practitioner, supported by action research, case study and interpretivist methodologies, data on the empirical manifestations of the environmental educational initiative were collected. Methods included semistructured interviews, focus groups, document analysis and participant observation. Findings indicate that schools and churches are important institutions that can positively influence patterns of identifying, knowledge about and agency for Boksburg Lake and can thus play a role in transforming hegemonic normalising ideologies. Important learning mechanisms identified included: Learning reflexively together within communities of practice that provide opportunities for active rather than passive learning; involving the youth as they are a group of people with notable enthusiasm, vision, energy and motivation; learning through information acquisition, investigation, action and deliberation; learning about abstract concepts and theoretical knowledge but embedding this in local realities; and learning that provides reference markers for how things can be different.

Mondi is an international packaging and paper company that manages over 300 000 ha of land in South Africa. After over a decade of working with Mondi to improve its wetland management, wetland sustainability practices were still not integrated into the broader forestry operations, despite some significant cases of successful wetland rehabilitation. An interventionist research project was therefore conducted to explore the factors inhibiting improved wetland management, and determine if and how expansive social learning processes could strengthen organisational learning and development to overcome these factors. In doing so, the research has investigated how informal adult learning supports organisational change to strengthen wetland and environmental sustainability practices, within a corporate plantation forestry context. How individual and/or group-based learning interactions translate to the collective, at the level of organisational change was a key issue probed in this study. The following three research questions were used to guide the research: 1. What tensions and contradictions exist in wetland management in a plantation forestry company? 2. Can expansive learning begin to address the tensions and contradictions that exist in wetland management in a plantation forestry company, for improved sustainability practices? 3. Can expansive social learning strengthen organisational learning and development, enabling Mondi to improve its wetland sustainability practices, and if so how does it do this? Cultural historical activity theory (CHAT) and the theory of expansive learning provided an epistemological framework for the research. The philosophy of critical realism gave ontological depth to the research, and contributed to a deeper understanding of CHAT and expansive learning. Critical realism was therefore used as a philosophy to underlabour the theoretical framework of the research. However CHAT and expansive learning could not provide the depth of detail required to explain how the expansive learning, organisational social change, and boundary crossings that are necessary for assembling the collective were taking place. Realist social theory (developed out of critical realism by Margaret Archer as an ontologically located theory of how and why social change occurs, or does not) supported the research to do this. The morphogenetic framework was used as a methodology for applying realist social theory. The expansive learning cycle was used as a methodology for applying CHAT and the theory of expansive learning; guiding the development of new knowledge creation required by Mondi staff to identify contradictions and associated tensions inhibiting wetland management, understand their root causes, and develop solutions. Through the expansive learning process, the tensions and contradictions become generative as a tool supporting expansive social learning, rather than as a means to an end where universal consensus was reached on how to circumvent the contradictions. The research was conducted in five phases: • Phase 1: Contextual profiling to identify and describe three activity systems in Mondi responsible for wetland management: 1) siviculture foresters; 2) environmental specialists; 3) community engagement facilitators. The data was generated and analysed through through document analysis, 17 interviews, 2nd generation CHAT analysis, and Critical Realist generative mechanism analysis; • Phase 2: Analysis and identification of tensions and contradictions through a first interventionist workshop. Modelling new solutions to deal with contractions, and examining and testing new models in and after the second interventionist workshop; • Phase 3: Implementing new models as wetland management projects and involved project implementation. This included boundary crossing practices of staff in the three activity systems, reflection and re-view in a further five progress review/interventionist workshops, and a management meeting and seminar; • Phase 4: Reflecting on the expansive learning process, results, and consolidation of changed practices, through nine reflective interviews and field observations; • Phase 5: Morphogenic/stasis analysis of the organisational change and development catalysed via the expansive social learning process (or not). The research found that expansive social learning processes supported organisational learning and development for improved wetland management by: 1) strengthening the scope, depth, and sophistication of participant understanding; 2) expanding the ways staff interact and collaboratively work together; 3) democratising decision making; 4) improving social relations between staff, reducing power differentials, and creating stronger relationships; 5) enhancing participant reflexivity through deeper understanding of social structures and cultural systems, and changing them to support improved wetland and environmental practice of staff, and developing the organisational structures and processes to strengthen organisational learning and development; and 6) using the contradictions identified as generative mechanisms to stimulate and catalyse organisational learning and development for changed wetland/environmental management.

This study addresses a core problem that was uncovered in records from coastal management monitoring initiatives on the eastern coast of Tanzania associated with the application and use of coastal monitoring indicators developed by external development partners for the coastal zone. These records suggest that local communities, who are key actors in participatory monitoring of coastal and marine resources, face many challenges associated with adapting and applying the said frameworks of indicators and monitoring plans. These indicators tend to be scientifically abstracted and methodologically reified; given prevailing contextual and socio‐cultural realities amongst them. The research project addresses the following key research question: How can processes of abstraction, conceptualisation, and representation of TEK contribute to the development of coastal management indicators that are less reified, more contextually and culturally congruent, and which may potentially be used by resource users in the wider social learning process of detecting trends, threats, changes and conditions of mangrove and fisheries resources? In response to the contextual problem and the research question, the study employs processes of abstraction and experiential learning techniques to unlock knowledge that local communities have, as an input for underlabouring existing scientific indicators on the Eastern coast of Tanzania. The research is constituted as critical realist case study research, involving two communities on the eastern coast of Tanzania, namely the Moa and the Boma communities (in Mkinga coastal district). Overall, the study involved 37 participants in a series of interviews, focus group discussions, and experiential learning processes using visualised data, and an experiential learning intervention workshop, and follow‐ups over a period of 3 years. The study worked with mangroves and fisheries to provide focus to the case study research and to allow for in‐depth engagement with the assumptions and processes associated with indicators development and use. Through the above mentioned data generation processes, critical realist analysis, and experiential learning processes involving abstraction and representation of traditional ecological knowledge held by mangrove restorers and fishers in the study areas, the study uncovers possible challenges of adapting and applying scientific indicators in participatory monitoring of a mangrove ecosystem. Using ampliative modes of inference for data analysis (induction, abduction and retroduction) and a critical realist scientific explanatory framework known as DRRREI(C) (Resolution, Re‐description, Retrodiction, Elimination, Identification, & Correction) the study suggests a new approach that may lead to the development of a framework of indicators that are less reified, more congruent to users (coastal communities), and likely to attract a wider context‐based social learning which favours epistemological access between scientific institutions (universities inclusive), and local communities. It attempts to establish an interface between knowledge that scientific institutions produce and the potential knowledge that exists in local contexts (traditional ecological knowledge), and seeks to widen and improve knowledge sharing and experiential learning practices that may potentially benefit coastal and marine resources in the study area. As mentioned above, the knowledge and abstraction processes related to the indicators development focussed on the mangrove ecosystem and associated fisheries, as engaged in the two participating communities in the eastern coast of Tanzania. The specific findings are therefore limited by the case boundaries, but the methodological process could be replicated and used elsewhere. The study’s contributions are theoretical and methodological, but also social and practice‐centred. The study brings into view the need to consider the contextual relevance of adapted knowledge, the capacity or ability of beneficiaries to adapt and apply scientific models, frameworks or tools, and the potential of local knowledge as an input for enhancing or improving monitoring of mangroves and mangrove‐based fisheries. Finally, the study comes up with a framework of indicators which is regarded by the coastal communities involved in the study as being less reified, more contextually and culturally congruent, and which may potentially be used in detecting environmental trends, threats, changes and conditions of mangrove and fisheries resources, and attract wider social learning processes.

Choisir un partenaire est une décision importante pour les organismes à reproduction sexuée. Une possibilité est de copier le choix d'autres individus, c'est ce que l'on appelle l'imitation du choix du partenaire, phénomène bien étudié chez les vertébrés, mais peu connu chez les invertébrés. Cette thèse porte sur l'imitation du choix du partenaire chez la drosophile (Drosophila melanogaster). J'ai tout d'abord montré que les femelles pouvaient être influencées dans leur choix par l'observation d'une seule autre femelle et que ceci était corrélé avec la pression atmosphérique. J'ai ensuite étudié si la préférence pour un type de mâle pouvait être transmise au sein de la population. Enfin j'ai commencé un travail exploratoire pour identifier les molécules impliquées dans ce processus d'imitation. Ces résultats montrent que les drosophiles peuvent exprimer des comportements complexes conduisant potentiellement à la transmission culturelle, l'isolement reproductif et la spéciation. / Mate choice is a major fitness-affecting decision in sexually reproducing organisms. A form of mate choice is mate copying, in which females choose potential mates by copying the mate choice of conspecifics. While many studies documented mate copying in vertebrates, little is known about this behaviour in invertebrates. In this thesis, I studied mate copying in Drosophila melanogaster females. I showed that female flies can build a sexual preference for one male characteristic after witnessing a single mate choice event and that the efficiency of mate copying correlates with air pressure and its variations. Then I studied the characteristics of mate copying to see whether a preference for one type of male can be transmitted into the population. Finally I tried to find some molecules that could be involved in this behaviour. These results indicate that fruit flies can express complex behaviour, which can potentially lead to cultural transmission, reproductive isolation and speciation.

Social intelligence is thought to have evolved as an adaptation to the complex situations group-living animals encounter in their daily lives. High levels of sociality provide individuals with opportunities to learn from one another. Social learning provides individuals with a relatively cheap and quick alternative to individual learning. This thesis investigated social learning in three corvid species: gregarious rooks (Corvus frugilegus) and jackdaws (Corvus monedula) and nongregarious, territorial Eurasian jays (Garrulus glandarius). In addition to that, the species' social structure was analysed and individual differences between members of each species were determined. Introducing the field of social learning research, I presented a new framework for investigating social learning, combining ecology, ethology and evolution. Experiments were conducted within that framework. I found that rooks and jackdaws develop social bonds and dominance hierarchies, whereas Eurasian jays do not. This is most likely related to their territoriality. In two experiments using two-action tasks, jackdaws learned socially. The underlying social learning mechanism was enhancement, which fits in with their feeding ecology. Rooks did not show social learning when presented with videos of conspecifics opening an apparatus. This might have been due to the difficulty of transferring information from videos or due to an ingrained 'affinity' to innovation and/or rapid trial-and-error learning overriding social learning processes. Individual differences along the bold/shy axis existed in all three species, but they were not stable across contexts. Thus, it seemed that the individuals perceived the two seemingly similar contexts that were designed to investigate neophobia and exploration (novel object in familiar environment; novel environment) as two different situations. The information may therefore have been processed by two distinct underlying mechanisms, which elicited different responses in each of the contexts. The implications of the findings of this thesis are discussed with regard to the new framework, integrating sociality, social learning and individual differences with the species' ecology.

M.A. / Conduct disorder is one of the most frequently diagnosed childhood disorders. The prevalence of this disorder has increased over the past few decades, which has ramifications for many facets of society, such as with families, justice systems, institutions involved with the rehabilitation of these children, and society as a whole. Conduct disorder has been found to be stable over time, and is therefore often associated with problems later in life such as violent crime, alcoholism, marital discord, and antisocial personality disorder. There have been many theories advanced for the aetiology of conduct disorder, but it is generally a bio-psychosocial model, rather than a single theory that receives the most attention when considering the development of this disorder. Included within such a model are variables such as a genetic component, neuropsychological factor, comorbid factor, socio-economic element, and a social learning component, that are involved with the development and maintenance of conduct disorder. One of the most consistently researched aspects involved within such models proposed for the aetiology of conduct disorder has been the role that certain parental styles have in the development of conduct disorder. Parental styles and the home environment have been consistently found to be a precursor of conduct disorder in foreign studies, however there is a lack of research within a South African context in this area. In order to establish whether there are specific styles of parenting related to conduct disorder in a South African sample, two measuring instruments were utilised, which were the Parental Bonding Instrument (PBI) and the Family Environment Scale (FES). The PBI and FES were administered to two groups of adolescents (n=80): one group of males (n=40) diagnosed with conduct disorder, and one group of males (n=40) without a diagnosis of conduct disorder. These instruments were used in order to establish if there were any differences in the perceived style of parenting between the two groups. The study yielded results that are similar to those found in foreign based studies. It was found with the South African sample, that a parenting style characterised by a low amount of care on the part of the mother, and overprotection on the part of the father was found within the conduct disorder group. Together these form a Parenting style of `affectionless control'. These parents were found to exert a high amount of control over their children, have a low expressiveness of emotions and feelings, have a low involvement with their children, and were poor at supervising and monitoring their children. These results indicate that parental styles could be a precursor of conduct disorder within a South African context. The implications of these results are discussed as well as the limitations of the study. Recommendations for future research and possible applications of the results are delineated.

A swarm intelligence system is a type of multiagent system with the following distinctive characteristics: (i) it is composed of a large number of agents, (ii) the agents that comprise the system are simple with respect to the complexity of the task the system is required to perform, (iii) its control relies on principles of decentralization and self-organization, and (iv) its constituent agents interact locally with one another and with their environment. <p><p>Interactions among agents, either direct or indirect through the environment in which they act, are fundamental for swarm intelligence to exist; however, there is a class of interactions, referred to as "interference", that actually blocks or hinders the agents' goal-seeking behavior. For example, competition for space may reduce the mobility of robots in a swarm robotics system, or misleading information may spread through the system in a particle swarm optimization algorithm. One of the most visible effects of interference in a swarm intelligence system is the reduction of its efficiency. In other words, interference increases the time required by the system to reach a desired state. Thus, interference is a fundamental problem which negatively affects the viability of the swarm intelligence approach for solving important, practical problems.<p><p>We propose a framework called "incremental social learning" (ISL) as a solution to the aforementioned problem. It consists of two elements: (i) a growing population of agents, and (ii) a social learning mechanism. Initially, a system under the control of ISL consists of a small population of agents. These agents interact with one another and with their environment for some time before new agents are added to the system according to a predefined schedule. When a new agent is about to be added, it learns socially from a subset of the agents that have been part of the system for some time, and that, as a consequence, may have gathered useful information. The implementation of the social learning mechanism is application-dependent, but the goal is to transfer knowledge from a set of experienced agents that are already in the environment to the newly added agent. The process continues until one of the following criteria is met: (i) the maximum number of agents is reached, (ii) the assigned task is finished, or (iii) the system performs as desired. Starting with a small number of agents reduces interference because it reduces the number of interactions within the system, and thus, fast progress toward the desired state may be achieved. By learning socially, newly added agents acquire knowledge about their environment without incurring the costs of acquiring that knowledge individually. As a result, ISL can make a swarm intelligence system reach a desired state more rapidly. <p><p>We have successfully applied ISL to two very different swarm intelligence systems. We applied ISL to particle swarm optimization algorithms. The results of this study demonstrate that ISL substantially improves the performance of these kinds of algorithms. In fact, two of the resulting algorithms are competitive with state-of-the-art algorithms in the field. The second system to which we applied ISL exploits a collective decision-making mechanism based on an opinion formation model. This mechanism is also one of the original contributions presented in this dissertation. A swarm robotics system under the control of the proposed mechanism allows robots to choose from a set of two actions the action that is fastest to execute. In this case, when only a small proportion of the swarm is able to concurrently execute the alternative actions, ISL substantially improves the system's performance. / Doctorat en Sciences de l'ingénieur / info:eu-repo/semantics/nonPublished

This dissertation consists of three journal articles that explored the effectiveness of a digital game, called EnerCities, in producing pro-environmental attitudes and behaviors by using a mixed-methods study approach. The first study was conducted as a quasi-experimental study among undergraduate students in the United States. Based on the Attitudinal Learning Instrument (ALI), this study found that the attitudinal learning gained from EnerCities influenced participants’ pro-environmental behavioral intentions significantly. This learning was retained until five weeks after game play according to the qualitative results of the study. The second study, conducted in India, used EnerCities to study the differences in attitudinal learning among high school students who played the game collaboratively or individually, using the ALI and Theory of Planned Behavior (TPB). Results showed that the attitudinal learning and its effect on pro-environmental behavioral intentions between collaborative and individual players was similar. This study also showed that EnerCities had significantly impacted the environmental attitudes and behaviors of the game players when compared to students who did not play any game, although all students had studied environmental studies through traditional instructional methods since elementary school. The third study, conducted among high school students in India, compared the environmental attitudes between game players and students who did not play any game based on the New Ecological Paradigm (NEP) scale. Both the unidimensional and multi-dimensional properties of the NEP were considered. It was found that EnerCities had impacted game players’ environmental attitudes significantly. All the three studies showed that digital games are more effective in promoting attitudinal (cognitive, affective, behavioral and social) learning compared to traditional instructional methods. This supports the implementation of digital games as a pedagogical tool in influencing environmental attitudes and behaviors.

Ce projet vise à déterminer si l’association présumée entre l’abandon des études chez les membres de la fratrie et le décrochage scolaire est plus ou moins forte en fonction de trois caractéristiques clés de la dyade, c’est-à-dire selon la correspondance des sexes (dyade de même sexe ou de sexe opposé), l’écart d’âge et le fait d’habiter ensemble ou non, à partir d’un échantillon d’adolescents canadiens (N= 545). Un tiers de l’échantillon est composé d’élèves ayant décroché de l’école secondaire, un second tiers d’élèves appariés ayant un profil semblable à celui des décrocheurs, mais ayant persévéré et d’un troisième tiers d’élèves normatifs. Les résultats montrent que les adolescents ayant un membre de la fratrie qui a décroché ont plus de chance de décrocher eux-mêmes que ceux dont la fratrie ne compte pas de décrocheurs. Cette association est plus forte lorsqu’il s’agit de dyades de mêmes sexes que de sexe opposé. De plus, les participants ayant un membre de leur fratrie décrocheur très proche en âge, c’est-à-dire avec un écart de deux ans et moins, ou assez proche en âge, c’est-à-dire avec un écart d’âge de trois à quatre ans, sont particulièrement à risque de décrocher, alors que les risques ne sont pas significativement plus élevés lorsque l’écart d’âge est plus grand, toujours comparativement à ceux sans décrocheurs au sein de la fratrie. Par contre, les résultats ne suggèrent aucun effet modérateur associé au partage ou non d’une résidence commune. Les implications des résultats pour la pratique et la recherche sont discutées. / The goal of this study is to determine whether the sibling dropout-adolescent dropout association depends on three key dyads’ characteristics, including: gender configuration (same-sex or opposite-sex dyad), age difference, and whether the adolescent and his or her dropout sibling live with in the same home or not. The data were collected from a sample of Canadian adolescents (N = 545). One-third of the sample consisted of students who had recently dropped out of high school, one-third were matched students who had a similar profile but persevered, and one-third were normative students not particularly at risk of dropping out. The results of the study show that adolescents with a sibling who has dropped out are more likely to drop out themselves than those with no siblings who have dropped out. This association is stronger when the dyads are of the same sex compared to opposite-sex dyads. In addition, participants who have a dropout sibling very close in age (with a gap of two years or less) or fairly close in age (with a gap of 3 to 4 years, are particularly at risk of dropping out, while the risks are not significantly higher when the age gap is larger, again compared to those without dropout siblings. However, the results do not suggest any moderating effects associated with whether or not a common residence is shared. The results’ implications for practice and research are discussed.

This study examined relationships among ethnic identity, gay identity, sexual sensation seeking, and HIV risk-taking behaviors among 302 men of color recruited from gay bars, bathhouses, community agencies, and the 1998 United States Conference on AIDS. The sample included 24% African American, 28% Latino, 25% Asian/ Pacific Islander, 19% Caucasian, 1% American Indian, and 3% other ethnicity. Logistic regression analysis identified sexual sensation seeking, having an undefined gay identity, being in a sexually exclusive relationship, not being HIV seronegative, and length of stay in the country (for those born overseas) as significant predictors of unprotected anal intercourse (insertive and penetrative) among men of color who have sex with men.