Information and punitiveness: trial reconstruction in Ireland

Purpose – The purpose of this paper is to report results from a rape trial reconstruction in Ireland. Design/methodology/approach – A studio audience of 100 members of the Irish public were selected to attend a TV programme by the Republic of Ireland’s national broadcasting organisation. This involved the examination of the sentencing of a rape case. The audience’s sentencing preferences were measured at the outset, when they had been given only summary information about the case, and later, when full details had been disclosed. Findings – Previous research examining changes in public attitudes to crime and punishment has shown that deliberation, including the provision of new information and discussion with others and experts, tends to decrease public punitiveness and increase public leniency towards sentencing. An experiment in Ireland, however, showed that providing information does not invariably and necessarily moderate punitive attitudes. This paper presents the results, and offers some explanations for the anomalous outcome. Research limitations/implications – The pre/post design, in which the audience served as their own controls, is a weak one, and participants may have responded to what they took to be the agenda of the producers. Due to the quality of the sample, the results may not be generalisable to the broader Irish population. Practical implications – Policy makers should recognise that the public is not uniformly punitive for all crimes. There is good research evidence to show that the apparent public appetite for tough punishment is illusory, and is a function of the way that polls measure public attitudes to punishment. Sentencers and those responsible for sentencing policy would benefit from a fuller understanding of the sorts of cases which illicit strong punitive responses from the public, and the reasons for this response. However any such understanding should not simply translate into responsiveness to the public’s punitive sentiments – where these exist. Innovative survey methods – like this experiment – which attempt to look beyond the top-of-the-head opinions by providing information and opportunities for deliberation should be welcomed and used more widely. Originality/value – There have been limited research studies which reports factors which may increase punitiveness through the provision of information and deliberation.

Pippi Longstocking, juvenile delinquent? Hachette, self-censorship and the moral reconstruction of postwar France

This paper looks at the blockages to the publication of children’s literature caused by the intellectual climate of the postwar era, through a case study of the editorial policy of Hachette, the largest publisher for children at this time. This period witnessed heightened tensions surrounding the social and humanitarian responsibilities of literature. Writers were blamed for having created a culture of defeatism, and collaborationist authors were punished harshly in the purges. In the case of children’s literature, the discourse on responsibility was made more urgent by the assumption that children were easily influenced by their reading material, and by the centrality of the young to the discourse on the moral reconstruction of France. As the politician and education reformer Gustave Monod put it: “penser l’avenir, c’est penser le sort des enfants et de la jeunesse.” These concerns led to the expansion of associations and publications dedicated to protecting children and promoting “good” reading matter for them, and, famously, to the 1949 law regulating publications for children, which banned the depiction of crime, debauchery and violence that might demoralise young readers. Using the testimonials of former employees, along with readers’ reports and editorial correspondence preserved in the Hachette archives, this paper will examine how individual editorial decisions and self-censorship strategies were shaped by the 1949 law with its attendant discourse of moral panic on children’s reading, and how national concerns for future citizens were balanced with commercial imperatives.

Evaluation of the realism of climate reconstruction using the Coexistence Approach with modern pollen samples from the Tibetan Plateau

We apply the Coexistence Approach (CoA) to reconstruct mean annual precipitation (MAP), mean annual temperature (MAT), mean temperature of thewarmestmonth (MTWA) and mean temperature of the coldest month (MTCO) at 44 pollen sites on the Qinghai–Tibetan Plateau. The modern climate ranges of the taxa are obtained (1) from county-level presence/absence data and (2) from data on the optimum and range of each taxon from Lu et al. (2011). The CoA based on the optimumand range data yields better predictions of observed climate parameters at the pollen sites than that based on the county-level data. The presence of arboreal pollen, most of which is derived fromoutside the region, distorts the reconstructions. More reliable reconstructions are obtained using only the non-arboreal component of the pollen assemblages. The root mean-squared error (RMSE) of the MAP reconstructions are smaller than the RMSE of MAT, MTWA and MTCO, suggesting that precipitation gradients are the most important control of vegetation distribution on the Qinghai–Tibetan Plateau. Our results show that CoA could be used to reconstruct past climates in this region, although in areas characterized by open vegetation the most reliable estimates will be obtained by excluding possible arboreal contaminants.

Tensor LRR and sparse coding-based subspace clustering

Subspace clustering groups a set of samples from a union of several linear subspaces into clusters, so that the samples in the same cluster are drawn from the same linear subspace. In the majority of the existing work on subspace clustering, clusters are built based on feature information, while sample correlations in their original spatial structure are simply ignored. Besides, original high-dimensional feature vector contains noisy/redundant information, and the time complexity grows exponentially with the number of dimensions. To address these issues, we propose a tensor low-rank representation (TLRR) and sparse coding-based (TLRRSC) subspace clustering method by simultaneously considering feature information and spatial structures. TLRR seeks the lowest rank representation over original spatial structures along all spatial directions. Sparse coding learns a dictionary along feature spaces, so that each sample can be represented by a few atoms of the learned dictionary. The affinity matrix used for spectral clustering is built from the joint similarities in both spatial and feature spaces. TLRRSC can well capture the global structure and inherent feature information of data, and provide a robust subspace segmentation from corrupted data. Experimental results on both synthetic and real-world data sets show that TLRRSC outperforms several established state-of-the-art methods.

Heterogeneous tensor decomposition for clustering via manifold optimization

Tensor clustering is an important tool that exploits intrinsically rich structures in real-world multiarray or Tensor datasets. Often in dealing with those datasets, standard practice is to use subspace clustering that is based on vectorizing multiarray data. However, vectorization of tensorial data does not exploit complete structure information. In this paper, we propose a subspace clustering algorithm without adopting any vectorization process. Our approach is based on a novel heterogeneous Tucker decomposition model taking into account cluster membership information. We propose a new clustering algorithm that alternates between different modes of the proposed heterogeneous tensor model. All but the last mode have closed-form updates. Updating the last mode reduces to optimizing over the multinomial manifold for which we investigate second order Riemannian geometry and propose a trust-region algorithm. Numerical experiments show that our proposed algorithm compete effectively with state-of-the-art clustering algorithms that are based on tensor factorization.