Mid Sweden University

The large-scale digitization of cultural heritage collections has intensified debates regarding the non-neutrality of metadata and the effects of descriptive regimes on the social construction of memory. This article critically examines the European project DE-BIAS (Detecting and Cur(at)ing Harmful Language in Cultural Heritage Collections), developed within the Europeana framework, discussing how its strategies contribute to mitigating historical and sociocultural biases in the representation of digital cultural heritage. The study draws on theoretical perspectives fromCritical Archival Studies and Information Science, which conceptualize archives and descriptive practices as arenas of power, and articulates these approaches with notions of epistemic justice, critical curation, and decolonial debates. Methodologically, this is a qualitative, exploratory, and critical study based on documentary analysis of the project's technical corpus, including reports, methodological guidelines, and institutional materials. The analysis focuses on three dimensions: (i) the multilingual vocabulary of potentially harmful terms and its bias typology; (ii) the automated harmful-language detection tool and its interpretative limitations; and (iii) the community engagement methodology oriented toward descriptive co-authorship. It is argued that de-biasing should be understood as an epistemological transformation and a practice of ethical governance, shifting the focus from terminological correction to institutional reflexivity, transparency, and social participation. The article also discusses the relevance of these approaches for the Global South, highlighting the need for critical appropriation and contextual adaptation in response to colonial legacies and infrastructural asymmetries. It concludes that initiatives such as DE-BIAS expand thepossibilities for more responsible and inclusive representation, provided that they are supported by long-term institutional policies, professional training, and accountability mechanisms.

The Draft-Based Iterative Feedback Accelerator (DIFA) is an AI-enhanced extension of the FAMS (Shahzad et al., 2023), which is designed to address challenges in providing scalable, high-quality formative feedback in higher education. Formative feedback is crucial for metacognitive development and self-regulated learning, yet its effective implementation is often limited in large cohorts, particularly for neurodiverse learners who need structured and transparent feedback. Moreover, the rise in students' reliance on generative AI raises concerns about deep learning, highlighting the need for assessment designs that promote meaningful cognitive engagement. DIFA transforms AI from a passive tool into an active pedagogical assistant within the feedback loop. It allows students to submit ongoing, incomplete work for evaluation, shifting focus from outcome-oriented grading to a process centered on reflection and continuous improvement. DIFA features reusable feedback fragments and a repository of common errors.

DIFA offers a modern approach to feedback by processing student drafts to generate context-specific responses that instructors can refine. This makes iterative feedback cycles more manageable within existing workloads. It enhances the Feedback and Assessment Management System (FAMS) by using natural language processing (NLP) to automate the categorization of feedback and create descriptive titles, improving organization and retrieval. When predefined feedback is unavailable, a lightweight language model trained on historical data generates relevant and clear responses, transforming the system into an adaptive feedback generator.

The architecture employs a teacher-in-the-loop approach where AI-generated feedback serves as a draft for instructors to review and refine. This ensures that AI enhances, rather than replaces, educators' professional judgment. By integrating AI into organizing and refining feedback, the DIFA system creates a continuous, data-driven feedback cycle that lightens instructors' workloads while delivering timely, personalized responses. Moreover, incorporating AI within a structured assessment framework helps prevent students' misuse of generative AI, positioning it as a tool for genuine learning.

Preliminary results show that this AI-enhanced architecture improves feedback delivery while maintaining the necessary depth and personalization for effective formative assessment. The DIFA system allows for iterative revisions rather than just final outputs, illustrating how explainable AI can enhance educational expertise. Future developments will include a student-facing interface that focuses on adaptive explanations and interactive feedback for neurodiverse learners

Research on digitalization and school leadership is emerging, but remainsfragmented and under-researched, as indicated in recent reviews. Based onthis, we conducted a scoping review of 38 empirical studies publishedbetween 2014 and 2024. The review focuses on the descriptive characteristicsof the field, how digitalization and digital school leadership are defined, whichthemes researchers address, and how existing research can be organized intoemerging research streams. The findings show that the field is expanding, butremains scattered across disciplines and journals, with a strong geographicalfocus on Europe, particularly Sweden, and parts of Asia. Although qualitativeand quantitative approaches are relatively balanced, mixed-method,longitudinal, and comparative designs are underused. As a result, much of theliterature provides snapshot accounts or examines the relationship betweendigital leadership and teachers’ technology use. Digital school leadership isoften treated as complex, yet this complexity is rarely unpacked analytically,leading to few studies of how leadership is enacted in everyday practice.Furthermore, the field reveals limited theoretical engagement and an unclearunderstanding of digitalization. The review also identifies a paradoxicalapproach to digital leadership. On the one hand, studies often lack explicitdefinitions; on the other hand, their research designs and methods create twoknowledge logics: a predefined, expectation-oriented logic and a situated,practice-oriented logic. Consequently, the review argues for research thatmoves beyond individualized understandings of digital leadership andexamines how it is enacted across organizational, technological, andgovernance contexts. It concludes by proposing a research agenda to addressthese gaps and guide future studies in more coherent and theoreticallygrounded directions

Acoustic imaging visualization is a core methodology in acoustics, enabling spatial analysis of sound sources and acoustic scenes. However, limited sensor availability in practical systems motivate approaches that enhance spatial resolution without increasing the hardware complexity. In this paper, we focus on upsampling virtually a tetrahedral 4-microphone array to a spherical 32-microphone array by estimating the covariance matrices of the channels employing deep learning techniques. Five neural network architectures are investigated for covariance upsampling for acoustic imaging using the real-world STARSS23 dataset. These models are developed to estimate a 32-microphone, time–frequency covariance matrix from a 4-microphone input covariance representation. The proposed architectures are based on 2D convolutional layers to capture the underlying spatial–spectral structure of covariance matrices, and are further enhanced with frequency dynamic convolution to model their frequency-dependent properties. The proposed architectures are evaluated in terms of root mean square error (RMSE) and using delay-and-sum beamforming acoustic imaging. Quantitative results show that all models outperform a random-guess baseline, which yields an RMSE of 0.548, with the best-performing architecture achieving an RMSE of 0.432. We analyze qualitatively the performance of the proposed models through beamforming heatmap visualizations derived from the 4-channel input covariance, the 32-channel ground truth, and the predicted 32-channel covariance matrices. These results demonstrate that covariance upsampling significantly enhances the effective performance of the 4-channel microphone array, producing sound maps that closely resemble those obtained with the 32-channel array.

Background: When a child has a life-threatening illness, guidelines emphasise the importance of providing holistic support to the whole family. Conversations about existential issues are essential in this support, yet staff often lack formal training in communication, particularly when addressing such issues as life and death. This qualitative study explored how paediatric healthcare professionals in Sweden experience and approachconversations about existential issues with children and their families facing life-threatening illness.

Methods: Eight professionals representing different disciplines and paediatric palliative care settings took part in individual interviews, analysed using an interpretative description approach.

Results: The results show that healthcare professionals experienced the existential conversations with families as emotionally intense, ethically complex, yet meaningful. A balance between courage, presence, sensitivity, and professional boundaries was crucial, as well as the need for collegial support. The importance of honest communication and age-appropriate ways of talking about illness and death was emphasized. They stressed that communication should be empathetic and culturally sensitive, without ready-made solutions, and based on the unique needs of the family. They also described organizational factors that influenced their ability to engage in existential conversations with the families.

Conclusions: Maintaining an open approach to existential conversations is essential to how these dialogues are carried out. Such an approach includes creating workplaces with an organizational culture that encourages existential conversations with patients and families. The importance of both personal and professional development is also emphasized, where collegial support and structures that promote professional growth, such as supervision and debriefing, strengthen the ability to address existential issues.

Background : Perseverative negative thinking is a key cognitive process associated with psychological distress in children and adolescents. Given Afghanistan’s ongoing challenges, a reliable self-report measure such as the PTQ-C is essential for accurate assessment among Afghan youth. Objective : The present study aimed to translate and validate the Dari/Farsi version of the Perseverative Thinking Questionnaire for Children (PTQ-C) in a sample of Afghan children and adolescents exposed to trauma. Method : The statistical population of this study consisted of 265 trauma-exposed Afghan children and adolescents. In this study, children and adolescents were selected using convenience sampling and completed an online survey (Qualtrics). To calculate test-retest reliability, 40 individuals were also selected and completed the PTQ-C within two weeks. In this study, the PTQ-C and the Depression Anxiety Stress Scale- 21 Items (DASS-21) were administered. Results : The confirmatory factor analysis (CFI) results showed that the single-factor model of PTQ-C had a good fit (χ2 /df = 2.51, CFI = .92, RMSEA = .07). Furthermore, PTQ-C showed a moderate positive association with depression, anxiety, and stress (r (265) = .34, 0.24, 0.32, p ' .05), an acceptable test-retest reliability, (r (40) = .72, p ' .05), and demonstrated good internal consistency (α = .83; ω=0.87). Conclusion : Given its strong psychometric properties, the Dari/Farsi version of the PTQ-C can be considered a reliable tool for both research purposes and clinical assessment. Its use is highly recommended for psychologists and mental health practitioners in diagnostic contexts as well as in investigative and applied studies. 

This chapter examines the development of filmmaking by Indigenous Sámi women in the Nordic countries in light of Sámi political activism during the 1970s, and contextualizes this movement in the context of later film and media policy. The chapter finds the lack of Sámi film productions between 1970 and 1979 to stand in stark contrast to Canada, where film served as a political medium for Indigenous women. The chapter traces a confluence of factors that have led to change, including the establishment of the International Sámi film Institute. Between 2009 and 2020, 76 Sámi films were produced, with 41 of them directed by women. Using the concept of visual sovereignty, the chapter concludes that there are no contemporary visual representations of the 1970s Sámi protests and cultural movements by Sámi women, but that recent audiovisual works has returned to represent some of these events. © 2026 selection and editorial matter, Anna Stenport, Maria Jansson, Mariah Larsson, and Scott MacKenzie; individual chapters, the contributors.

The efficient deployment of deep neural networks (DNNs) on resource-constrained edge devices necessitates a delicate balance between model accuracy, communication bandwidth, and computational latency. While existing methods often address partitioning, quantization, or compression in isolation, this paper introduces CO-NAS (Joint Neural Architecture Search for Compression, Quantization, and Edge-Cloud Model Offloading), a unified framework that jointly optimizes these three critical dimensions for collaborative edge-cloud inference. By utilizing an adapted version of the stochastic architecture search (SNAS) paradigm, CO-NAS leverages Temperature-Annealed Softmax relaxations to enable gradient-based optimization over discrete choices of partition points, compression ratios, and activation bit-widths. Extensive experiments on ResNet-50 and VGG-16 across CIFAR-100 and Fashion-MNIST datasets demonstrate the framework's efficacy. CO-NAS discovers Pareto-optimal configurations, such as a ResNet-50 model for CIFAR-100 achieving 70.1% accuracy with extreme compression (48× channel reduction and 1-bit quantization), representing a total compression of over 1500× on transmitted features. Crucially, exhaustive grid search validation confirms that CO-NAS reliably identifies top-performing architectures within the vast search space, while achieving a 21× -27× reduction in search cost compared to the exhaustive baseline. The results validate CO-NAS as a scalable and efficient approach for automating the co-design of accurate and communication-efficient neural networks for the edge-cloud paradigm. 

As pointed out by Floridi (2024) and Youvan (2025), the current AI hype shows several characteristics that resemble earlier tech bubbles such as the Dot-Com Bubble and the Cryptocurrency Bubble. These two bubbles were also centred around a technology with a potential to revolutionise multiple sectors, and at the same time had a tendency to cause tunnel vision. The general pressure to meet high expectations, contradicted by reports on technological challenges and AI slop (Ansari, 2025), raise concerns about sustainability. If the AI bubble bursts there could be drastic consequences across tech industry, job market, and the general economy (Youvan, 2025). Less has been published about how a bursting AI bubble might have a strong impact on different educational contexts. This study has the aim of analysing and discussing how the fields of higher education and lifelong learning would be affected by an AI bubble burst.

In the field of higher education most of the identified challenges related to GenAI will probably remain such as the need for rethinking assessment and course design (Perkins et al., 2025). There is still a need for concrete guidelines on how educators and student ethically might use generative AI (GenAI) in teaching and learning activities. Regarding lifelong learning, the need for an increase of lifelong learning has increased in modern society, creating new challenges for higher education to achieve lifelong learning. As lifelong learning includes learning throughout the entire life span, it can be seen as a way to strengthen individuals' opportunities to participate actively in society and to develop democratic and personal competencies. Here, the iterative improvement of GenAI tools could support flexibility individualisation, accessibility and individualisation for students. Further, lifelong learning is often emphasised in relation to the labour market's need for a well-educated and competent workforce (Jaldemark; 2021; OECD, 2021) of which competences in tools will be necessary in the future.  

If the AI bubble bursts, it may result in an interrupted development of GenAI tools. The existing toolbox works well to at least partly solve many types of traditional higher education assignments for students. The quality of speech synthesis, audio to text conversion, and tailored instructions are today much higher than just a few years ago.  Students’ use of chatbots as artificial tutors and study buddies will not be ended by a burst AI bubble.  At the same time, the iterative improvement of GenAI tools has reached a level of quality that could meet the high expectations on GenAI regarding individualisation and accessibility for students in higher education and therefore lifelong learning. 

The rapid advancement of generative artificial intelligence has significantly increased the sophistication, accessibility, and impact of deepfakes, creating urgent challenges for cybersecurity, digital trust, and information integrity. As synthetic media becomes increasingly indistinguishable from authentic content, organizations face growing risks related to misinformation, fraud, and identity manipulation.

This research explores emerging approaches to deepfake detection by integrating facial analytics, computer vision, and biometric-based artificial intelligence models. It introduces a multimodal detection framework that combines spatial–temporal facial analysis, audio–visual consistency validation, and embedding-based anomaly detection to identify inconsistencies in synthetic media. The approach emphasizes resilience against evolving adversarial techniques and highlights the importance of moving beyond single-modality detection methods.

In addition to technical considerations, the session examines key challenges, including dataset bias, model generalization, and the ongoing arms race between generative and detection technologies. The presentation also addresses governance and operational implications, focusing on how organizations can integrate deepfake detection into cybersecurity workflows while maintaining ethical and trustworthy use of artificial intelligence.

The discussion contributes to broader conversations on balancing innovation with safeguards, positioning deepfake detection as a critical capability within the evolving landscape of AI opportunities and challenges. Attendees will gain practical insights into current detection strategies, limitations, and future directions for securing digital environments against synthetic media threats.

The proliferation of deepfakes, enabled by advances in generative artificial intelligence, presents a growing challenge to communication ecosystems, democratic processes, and public trust in digital information. This study examines deepfakes through a communications and policy lens, focusing on how synthetic media disrupts information integrity, amplifies dis/misinformation, and complicates the verification practices of journalists, institutions, and the public. Concurrently, emerging AI-based detection tools have been developed to distinguish synthetic from authentic content—yet this remains a dynamic and unsettled contest. The race between generative and detection technologies is ongoing, and deepfakes have evolved into a multimodal phenomenon that has outpaced existing governance structures. While technical detection methods continue to evolve, gaps remain in governance frameworks, regulatory alignment, and institutional readiness to address the societal impacts of deepfakes.

Drawing on interdisciplinary insights from a scoping review spanning communication studies, information policy, and cybersecurity, this study analyses prevailing approaches to deepfake mitigation and surfaces critical deficiencies in transparency, accountability, and cross-sector coordination. From a comparative global perspective, regulatory responses vary considerably: the California AI Transparency Act and the EU AI Act represent substantive—if imperfect—policy efforts, yet much of the world remains without meaningful guidance or enforceable standards.

Authors propose a policy-oriented framework emphasizing three pillars: (1) standardized disclosure and labeling mechanisms for synthetic media, (2) institutional capacity-building for verification and media literacy, and (3) collaborative governance models involving public agencies, technology platforms, and academia. Moreover, the study has explored the ethical implications of detection technologies, including risks of overreach, bias, and unintended censorship. By situating deepfake detection within broader communication policy debates, this research contributes to a more holistic understanding of how societies can balance innovation with the protection of information integrity and public trust.

The rapid development of generative artificial intelligence (GenAI) during recent years has opened new challenges and opportunities for technology enhanced learning in higher education. In the field of artificial intelligence in education (AIED), there has been frequent discussions on AI cheating and how to redesign assessment. This study has explored the opportunities of involving GenAI in learning activities with a chatbot as a member in students' group work. How would that affect group dynamics and which chatbot roles and patterns can be identified in the conversations between students and chatbots in the solving of an assignment on crisis management? Data have been collected from 10 group works with an assignment that involved role play dialogues with a student chosen GenAI bot. All dialogues with the GenAI bots have been logged as appendices to the solved assignment. The Conversation analysis approach was used to find patterns, strategies and organisation of interaction in logged conversations.

Findings show that students actively structured and directed the interaction with the GenAI chatbot through recurring strategies. Students framed the task and positioned the chatbot as a contributor, while also redefining the interaction by introducing roles that transformed the dialogue into a simulated professional context. Although this was a part of the instruction within the course, the students further expanded the interaction by adopting roles themselves. They used interview-like questioning to elicit assessments and expert-oriented responses, and guided the dialogue through a stepwise progression from scenario construction to situational analysis and concrete actions. Overall, the students maintained control of the interaction and used the chatbot strategically to advance the problem-solving process. Another pattern observed was how students positioned the chatbot differently in the interaction. At times the chatbot was addressed as a "you", positioning it as an external expert, while in other instances students referred to it as part of a collective "we", treating it as a member of the problem-solving team. 

This study shows that students do not simply use GenAI chatbots as information sources, but also actively structure and manage the interaction through different conversational strategies by positioning the chatbot as both an expert and a collaborative participant in the problem-solving process. The findings suggest that integrating GenAI into group work can support structured exploration of complex problems, while also highlighting how students strategically shape the interaction to align the chatbot with the goals of the assignment. Further research is needed to explore how task designs influence students' interaction strategies when working with GenAI in collaborative learning activities. Finally, it would be interesting to compare the quality of group assignments solved with bot members, with the ones solved by humans only. 

The structural integrity of composite rocket propellants is critical for the reliability of rocket motors, as insufficient integrity can lead to failure or even motor explosion. However, the material properties of such propellants degrade over time. Ageing is typically monitored using tensile testing, a time-consuming and destructive method that requires substantial sample volumes. Consequently, alternative characterisation techniques are sought. One promising approach is dynamic mechanical analysis (DMA), which requires samples that are orders of magnitude smaller than those used for tensile testing. The aim of this study was to establish relationships between strain and elastic modulus (from tensile testing), storage modulus (from DMA), and hardness. Composite propellants based on hydroxyl-terminated polybutadiene, ammonium perchlorate, and aluminium were manufactured, subjected to accelerated ageing at 75–105 °C in an inert atmosphere for up to 190 days, and characterised by tensile testing, DMA, and hardness measurements. For each ageing temperature, the relative change of each property was plotted as a function of time. Empirical expressions linking strain, elastic modulus, storage modulus, and hardness were derived from the Arrhenius equation. These empirical models enable reliable conversion between, for example, storage modulus and strain for lifetime prediction, thereby demonstrating that DMA can serve as a viable alternative or complement to tensile testing. 

Students' declining interest in school science as they transition from primary to secondary education has raised concerns about how science education connects to students' lives and their beliefs about science. This study examines ninth-grade students' attitudes towards school science through two value-related dimensions: perceived relevance and trust in science. Drawing on Expectancy-Value Theory (EVT), we empirically examine students' interest in school science and investigate its relationship with their perceptions of relevance (utility and attainment value) and trust in science as a problem-solving resource. A convergent mixed-methods case-study design was employed with 63 students (age 15) in one Swedish lower secondary school. Quantitative data were collected using selected items from the ROSES survey and analysed through descriptive statistics and separate regression models for relevance and trust. Qualitative data consisted of students' argumentative texts responding to paired statements on relevance and trust, which were analysed using thematic analysis. Findings indicate that perceived relevance has the strongest association with interest, while trust is positively related but insufficient on its own. Qualitative themes illustrate how students experience science as personally meaningful, societally important, or abstract and distant. The study discusses implications for designing science teaching that supports both perceived relevance and informed trust.

Aim: To investigate midwifery students’ interests and views towards MCoC models and to validate an instrument measuring these views.

Background: Midwifery continuity of care (MCoC) is widely recognised for its benefits to maternal and neonatal health. Despite this, access to MCoC models remains limited in Sweden and midwifery students receive minimal clinical exposure to this model of care.

Design: Mixed-methods cross-sectional study Methods: Data were collected from a national sample of 115 midwifery students in Sweden using an online questionnaire between 2022 and 2023. Descriptive statistics were used to present students' characteristics and interests in MCoC. Questions about students’ views of MCoC models were subjected to a principal component analysis (PCA). For associations between PCA components and students’ socio-demographic characteristics, Analysis of Variance was performed. Free-text responses were thematically analysed.

Results: Midwifery students held strong and positive views about the importance of MCoC and many were interested in working in such a model. However, some expressed concerns about work-life balance. The PCA resulted in two components: Relational aspects and Practical aspects, with Cronbach's alpha values of 0.735 and 0.751, respectively. The free-text responses confirmed the content of the components, with the themes Developing confidence in the professional role, Enhancing midwifery skills and expertise, Stronger relational bond and Challenging work conditions. Few socio-demographic characteristics were associated with the components.

Conclusions: Working within a MCoC model was generally perceived as appealing by the midwifery students. Authorities and educational institutions must facilitate providing education and opportunities for MCoC models in Sweden. 

This work investigates the coupled influence of switching phase shift and external voltage amplitude in the Synchronized Switch Damping on Voltage (SSDV) method on the modal characteristics of a piezoelectric structure. A lumped-parameter electromechanical model was formulated and validated using a hybrid energy harvester setup. Numerical analyses performed over a broad parameter space reveal that the phase shift—traditionally viewed as a performance-degrading factor—can be exploited as a critical control variable for simultaneous damping and stiffness modulation. The findings demonstrate that specific phase–voltage combinations can amplify the achievable resonance frequency shift by up to a factor of two compared to conventional tuning-only strategies. Furthermore, experimental results corroborate that optimized phase-shifted configurations significantly outperform standard zero-phase SSDV. Specifically, for voltage coefficients ku of 0.3, 0.6, and 0.7, the proposed approach achieved a reduction in the area under the frequency response functions (FRFs) by 15%, 17%, and 22%, respectively. These results highlight the potential of phase-shifted SSDV for high-precision applications, such as the stabilization of machining processes, where the dual-action control of damping and structural resonance is essential for enhancing operational stability and surface finish. 

Wildfire risk at the Wildland–Urban Interface (WUI) emerges through everyday landscape features; features cut across areas of responsibility and impact on diverse land-uses. In Sweden, wildfire hazard is shaped primarily by human activity and rapidly drying vegetation, yet public awareness remains low in part due to the historical rarity of large fires. Recent extreme fire events in 2014 and 2018 have however exposed a growing vulnerability, particularly within dispersed settlements embedded in forested terrain. In this paper we examine perceptions of wildfire mitigation and adaptation through three scenario-based workshops conducted in Oskarshamn, Sundsvall, and Gothenburg. The scenario narratives, grounded in real fire events, were used to facilitate discussion, draw out local knowledge, and examine assumptions of fire risk, responsibility, and mitigation actions. Analysis of the workshop discussions has been structured around five mitigation strategies: vegetation management, structure and garden hardening, community hardening, education, and enforcement and incentives. The analysis reveals both shared concerns and need to recognise context across regions. Participants emphasized the importance of vegetation management, diversified forest structure, improved building and garden practices, and place-specific education, while highlighting persistent challenges linked to unclear responsibilities, economic constraints, and limited regulatory support. Our study demonstrates the usefulness of scenario workshops for collective learning and realising local capacities, for living with fire risk in the Swedish WUI.

In mountainous landscapes, where visitor numbers are increasing, it is essential to balance biodiversity conservation, indigenous practices, and visitor experiences through planning and management. Our study investigates visitor experiences, attitudes, and awareness of ecological impacts in the Jämtland mountains, Sweden. In summer 2021, we conducted two qualitative landscape observations, a quantitative study, collecting 253 contact cards, and conducting 13 semi-structured interviews with hikers. Findings revealed that wear and tear was the most prevalent issue, followed by overcrowding. Despite concerns about wildlife impacts, reindeer sightings were anticipated and cherished, seen as integral to the visitor experience. Concerns were raised regarding the Right of Public Access. Opinions on management tools like zoning restrictions were mixed, with support for monitoring but concern about implications. Views on visitation fees varied, emphasising equitable access and voluntary contributions. Education and information were viewed positively, as was the importance of greater understanding of Sami culture. The results underscore the need for further studies on synergies among land-use interests to support multifunctional landscapes that benefit biodiversity and society while recognising recreation’s positive impacts. In this context, the Visitor Use Management framework may offer a practical approach to managing this socio-ecological system of multiple values.