Stellarator boundary optimisation faces a fundamental numerical challenge: the extreme disparity between low- and high-mode amplitudes creates an optimisation landscape in which direct full-spectrum approaches typically converge to poor local minima. Traditionally, this challenge has been addressed through a computationally expensive, multi-step Fourier continuation, in which low Fourier modes are optimised first, followed by the gradual incorporation of higher modes. We present exponential spectral scaling (ESS), a technique that applies a mode-dependent exponential scaling factor to each Fourier mode. Our primary implementation uses the $L_{\infty }$ norm to determine the scaling pattern, creating a square spectral decay profile that effectively reduces the dynamic range of optimisation variables from 6–7 orders of magnitude to 2–3. This scaling aligns with the natural spectral decay of physically meaningful configurations and enables direct single-step optimisation using the full spectrum of boundary Fourier modes. ESS eliminates arbitrary staging decisions and reduces computation time by a factor of ${\sim}2{-}5$ in benchmark cases. In addition to accelerating optimisation, ESS improves robustness, reducing sensitivity to initial conditions and increasing confidence in avoiding local optima. We demonstrate the effectiveness of ESS across both quasi-axisymmetric (QA) and quasi-helically symmetric (QH) configurations, using two distinct optimisation toolkits: simsopt and desc.

For many years, political scientists have debated over voter competence in direct democracy. At the core of the discussion is whether this central institution enlightens citizens about political facts. However, scholars have primarily examined if direct democracy fosters general political knowledge even though referendums and ballot initiatives are policy-specific in nature, as citizens vote on particular political proposals. By utilising a range of unique panel survey data collected around four Danish European Union referendums, I show that voters’ knowledge of policy-specific information markedly increased during the campaigns. I also combine the survey data with an original media content analysis and find that the learning of issue-specific facts is more related to the opportunities provided by the media information environment than to individual ability or motivation. These results suggest that a broad group of voters acquire policy-specific facts that help them make informed choices when they are granted full control of political decision-making.

This manifesto was originally submitted into and went on to win the senior category in the 2025 Classical Association ‘Write | Speak | Design’ competition. Through a mix of research and my personal classics journey, it argues that the contemporary relevance and remote accessibility of classical subjects, along with the academic joy they encourage in learners, make the study of the ancient past as important as ever and highlights ways these areas could be harnessed to increase the importance of classics further.

This paper reviews research from the field of language teaching into post observation feedback i.e. the discussion that takes place after an observer has watched a pre-service or in-service teacher’s lesson. Post observation feedback is discussed with reference to four main themes: (1) perceptions of feedback; (2) reflection; (3) relationships (with two sub-themes of identity and facework); and (4) observer training. This review indicates that while the fields of language teaching and applied linguistics are leading research into post observation feedback, there remains important and interesting avenues for future research, which are discussed in this paper.

With the growing application of artificial intelligence (AI) and machine learning (ML), great potential exists to leverage these technologies in paleontology. Relative to many other scientific fields, a challenge of ML applied to paleontology is small sample sizes, particularly for fossil vertebrates. Shark teeth, abundant in the fossil record, provide a model system to use ML across varying sample sizes. Here we use six classes (taxa) of Neogene shark teeth for taxonomic identification, including a curated dataset of 3150 images. Each class was evaluated using an 80% training and 20% validation split, with a separate, external test set of 25 samples per class. Pretrained models perform well (accuracy > 90%), providing a strong baseline for classification. However, enabling fine-tuning of the ML model to identify fossil shark teeth improves performance considerably. Likewise, sample size per class also affects the accuracy of the models’ classifications. Smaller sample sizes (n = 50 individuals per class) yielded a mean accuracy of 93.4%, but plateaued at ~99% between 200 and 500 images per class. Confidence likewise increases with larger samples, from 81.8% (n = 50 individuals per class) to >90% (n = 300 to 500 individuals per class). Misidentifications followed consistent patterns, reflecting morphological similarities and/or poor preservation. Artificially increasing the training datasets using data augmentation improves the confidence of identifications. This research indicates that relatively small samples of vertebrate species (~50 to 500 individuals per class) can effectively train an ML model to identify these shark teeth with high levels of accuracy.

This essay explores the conceptual and methodological contribution of a spatial understanding of labour law, examining the ways in which labour laws create sites of inclusion and exclusion that can be subverted by worker action. It argues that labour relations cannot be apprehended without considering their place in space. It further argues that labour laws tend to foster inertia within industrial relations by recognizing certain workspaces while failing to adapt to the dynamic geographies of the workplace. Methodologically, this implies a shift from a neutral discourse of rights to one that is anchored in social life where workers converge. This essay suggests that recognizing concrete and dynamic spaces of labour within legislation can lead to upholding diverse voices at work, especially from workers traditionally left in the margins, like women, minorities, and migrants.

Digital technologies provide a novel environment for political activities and, more specifically, for interactions between citizens and political actors. The scholarly literature on these topics is flourishing. On the one hand, studies of political communication emphasise the changing nature of election campaigns and the reshaped relationship between leaders and supporters. On the other hand, the literature on political parties examines the organisational implications of such a digital shift in more detail. Against this backdrop, this study investigates the opinions and participatory attitudes of party members towards the new digital participation opportunities that their party organisations provide. To do so, we rely on original individual survey data. More specifically, we will use data derived from a survey administered to Partito Democratico members in Italy at the beginning of 2022 (approximately 4000 respondents). Precisely, we aim to identify the profiles of party members according to their (degree of) digital activities by controlling for variables such as length of membership, levels of intraparty activism, and evaluation of intraparty democracy. Moreover, we investigate the changing relationship between members and their party organisations in the new digital ecosystem.

We examine how the presence of active galactic nuclei (AGN) correlates with location in large-scale cosmic structures using the Galaxy and Mass Assembly (GAMA) survey across the G09, G12, and G15 fields. Our sample contains 18 927, 9 273, and 1 148 galaxies for highly dense filaments, moderately dense tendrils, and highly underdense voids, respectively. AGN are identified among emission-line galaxies using Baldwin-Phillips-Terlevich (BPT) diagnostic diagrams based on [NII], [SII], and [OI]. We compare AGN fractions across filament, tendril, and void regions and as a function of distance from the nearest filament centreline. Our results reveal a mild excess in filaments compared to voids when using [SII]- and [NII]-based classifications, while no significant environmental dependence is found for [OI]-based classifications. Overall, we find a weak environmental trend with AGN activity, which suggests that the local environment does not always dominate AGN activity; instead, secular processes are likely to be at play. Our findings are consistent with previous studies reporting only marginal preferences for overdense environments for AGN.

Snow cover influences sea ice thermodynamics and mass balance, making its distribution and properties critical to polar research. Grounded icebergs in coastal Antarctica substantially affect surface snow distribution and landfast sea ice patterns, which have received limited scientific attention. To address this gap, this study integrates airborne laser scanning observations with numerical snow transport simulations to investigate snow distribution on landfast ice around icebergs, emphasizing the influence of wind and iceberg size. Observations show that persistent wind directions shape characteristic snow patterns around icebergs, with substantial windward and lateral drifts and an elongated snow-depleted region in the lee. Data further reveal that snowdrift size scales nonlinearly with iceberg size, indicating reduced snow accumulation efficiency for larger icebergs, which simulations partially captured. This study also highlights the key role of wind direction shifts in reproducing measured snow distributions and suggests that the maximum extent of snowdrifts is constrained by peak wind speeds encountered on site. Together, our findings show that iceberg-induced snowdrifts connect ice shelf and fast ice dynamics, reflect local wind conditions and provide key insights into snow mass balance on Antarctic landfast sea ice.

According to the Roy–Borjas model, the most talented workers will choose to migrate to countries exhibiting high income inequalities to achieve the highest earnings. The purpose of this article is to highlight that income inequalities in the country of origin, particularly the nature of inequalities, will affect high-skilled emigration. If the home country rewards productive efforts and sanctions unproductive behaviours (such as rent-seeking), emigration declines. We test this hypothesis by relying on panel data of 30 OECD countries for the period from 1990 to 2010. Two econometric techniques are used: the ordinary least squares and the system-Generalized Method of Moments estimation to tackle the endogeneity issue. The results show that when income inequalities in the home country are conditioned by the institutions’ quality, there is a negative relationship between inequalities and high-skilled emigration, suggesting that productive inequalities are detrimental to emigration. Thus, developed countries facing high-skilled emigration must change the nature of inequalities by reforming their institutions in order to attract and retain the most talented workers. Implementing institutions that reward productive efforts would limit high-skilled emigration.

In a puzzling passage from his computistical handbook, Byrhtferth of Ramsey asks his students to imagine the Venerable Bede sitting in Moses’ tabernacle and teaching them about the calculation of time. This article considers how Bede was portrayed as a mediator of divine wisdom about computus in Byrhtferth’s Enchiridion and Epilogus, culminating in Bede’s elevation to the eternal space of the tabernacle. Building on Mary Carruthers’ ‘machines of meditation’ and Faith Wallis’ work on Byrhtferth’s diagrams as ‘visual exegesis’, it argues that a collection of riddling references to the tabernacle across Byrhtferth’s canon amount to a cosmography in which the tabernacle is conceived as an exegetical model for God’s presence in time and space.

This symposium grew out of dissatisfaction with the existing theories of institutions. Notwithstanding significant progress in the analysis of the macro-institutions through which systemic rules and norms are established and the micro-institutions through which actors decide and implement transactions within the playing field thus defined, researchers working along one or the other dimension faced a critical and largely unanswered question: how to bridge the gap between these two institutional layers? The selected articles assembled in this issue came out of efforts to identify and understand within a unified theoretical framework the arrangements through which these layers interact. Building on contributions in economics and other social sciences as well as from in-depth empirical studies, these articles explore the relevance of the concept of ‘meso-institutions’ to designate and characterize the devices (e.g. regulatory agencies) and mechanisms (e.g. guidelines) that connect the macro- and micro-institutional layers.

Regardless of whether one is analyzing quantitative data from research involving generative artificial intelligence (GenAI) or more classical methods, testing for normality remains a necessary step in statistical analysis. Although over 60 methods have been proposed for assessing univariate normality, previous systematic reviews show that normality testing remains underreported in L2 research. This paper addresses this gap by first reviewing the concept of normality and its role in parametric statistical inference. We then examine 12 normality assessment methods including five graphical and seven analytical methods selected based on their prominence in statistical literature and availability in commonly used software. Each method is explained in terms of its underlying mechanism and sensitivity to specific forms of nonnormality, such as skewness, tail heaviness, and multimodality. In the second part of the study, we review 237 empirical articles published between 2020 and 2025 in ten selected L2-focused Q1 journals, using AI-assisted annotation. Our findings reveal inconsistencies in how graphical tests are reported, a tendency to rely on tests such as the Kolmogorov-Smirnov without explicit attention to sample size constraints, and limited justification provided for critical values of skewness and kurtosis. These results indicate some divergence between recommended statistical practices and the procedures for normality testing reported in the L2 publications examined. The paper concludes with actionable recommendations for selecting and interpreting normality tests in L2 research contexts.