In recent years, scholars have devoted a great deal of attention to the history of scholarship in general and, more specifically, to the emergence of critical historical and anthropological literature from and within ecclesiastical scholarship. However, few studies have discussed the Jewish figures who took part in this process. This paper analyzes the role played by historiographical and ethnographical writing in seventeenth- and eighteenth-century Italian Jewish–Christian polemics. Tracing various Christian polemical ethnographical depictions of the Jewish rite of shaking the lulav (sacramental palm leaves used by Jews during the festival of Sukkot), it discusses the variety of ways in which Jewish scholars responded to these depictions or circumvented them. These responses reflect the Jewish scholars’ familiarity with prevailing contemporary scholarship and the key role of translation and cultural transfers in their own attempts to create parallel works. Furthermore, this paper presents new Jewish polemical manuscript material within the relevant contexts, examines Jewish attempts to compose polemical and apologetic ethnographies, and argues that Jewish engagement with critical scholarship began earlier than scholars of this period usually suggest
In recent years, pre-modern beds have generated extensive scholarly interest. Their social, religious, and economic importance has been rightfully highlighted in the study of domestic piety. Yet, concern has primarily focused on beds in late medieval English homes. This essay uses Hebrew texts from thirteenth-century Southern Germany, primarily Sefer Hasidim, to further this analysis of the role of the bed in shaping medieval domestic devotion. Jewish notions about the social, moral, and sexual significance of the bed reflect those identified in late medieval Christian culture. These ideas inspired numerous rituals practiced in Jewish homes. Yet, the bed and the remnants of sex assumed to be found in it also frustrated Jewish attempts to perform domestic devotion. These findings highlight the complicated nature of the home and how medieval people had to navigate both its opportunities and challenges in order to foster a rich culture of domestic devotion.
In recent decades the relationship between tantric traditions of Buddhism and Śaivism has been the subject of sustained scholarly enquiry. This article looks at a specific aspect of this relationship, that between Buddhist and Śaiva traditions of practitioners of physical yoga, which came to be categorised in Sanskrit texts as haṭhayoga. Taking as its starting point the recent identification as Buddhist of the c.11th-century Amṛtasiddhi, which is the earliest text to teach any of the methods of haṭhayoga and whose teachings are found in many subsequent non-Buddhist works, the article draws on a range of textual and material sources to identify the Konkan site of Kadri as a key location for the transition from Buddhist to Nāth Śaiva haṭhayoga traditions, and proposes that this transition may provide a model for how Buddhist teachings survived elsewhere in India after Buddhism’s demise there as a formal religion.
Publisher: Association for Computational Linguistics
Country: United Kingdom
Project: EC | SUMMA (688139), EC | SUMMA (688139)
In this paper we introduce a new publicly available dataset for verification against textual sources, FEVER: Fact Extraction and VERification. It consists of 185,445 claims generated by altering sentences extracted from Wikipedia and subsequently verified without knowledge of the sentence they were derived from. The claims are classified as Supported, Refuted or NotEnoughInfo by annotators achieving 0.6841 in Fleiss $\kappa$. For the first two classes, the annotators also recorded the sentence(s) forming the necessary evidence for their judgment. To characterize the challenge of the dataset presented, we develop a pipeline approach and compare it to suitably designed oracles. The best accuracy we achieve on labeling a claim accompanied by the correct evidence is 31.87%, while if we ignore the evidence we achieve 50.91%. Thus we believe that FEVER is a challenging testbed that will help stimulate progress on claim verification against textual sources. Comment: Updated version of NAACL2018 paper. Data is released on http://fever.ai
Indic rites of purification aim to negate the law of karma by removing the residues of malignant past actions from their patrons. This principle is exemplified in the Kahika Mela, a rarely studied religious festival of the West Himalayan highlands (Himachal Pradesh, India), wherein a ritual specialist assumes karmic residues from large publics and then sacrificed to their presiding deity. British officials who had ‘discovered’ this purificatory rite at the turn of the twentieth century interpreted it as a variant of the universal ‘scapegoat’ rituals that were then being popularized by James Frazer and found it loosely connected to ancient Tantric practises. The However, observing a recent performance of the ritual significantly complicated this view. This paper proposes a novel reading of the Kahika Mela through the prism of karmic transference. Tracing the path of karmas from participants to ritual specialist and beyond, it delineates the logic behind the rite, revealing that the culminating act of human sacrifice is, in fact, secondary to the mysterious force that impels its acceptance.
Publisher: Association for Computational Linguistics (ACL)
Country: United Kingdom
Project: EC | SUMMA (688139), EC | QT21 (645452)
Analysing translation quality in regards to specific linguistic phenomena has historically been difficult and time-consuming. Neural machine translation has the attractive property that it can produce scores for arbitrary translations, and we propose a novel method to assess how well NMT systems model specific linguistic phenomena such as agreement over long distances, the production of novel words, and the faithful translation of polarity. The core idea is that we measure whether a reference translation is more probable under a NMT model than a contrastive translation which introduces a specific type of error. We present LingEval97, a large-scale data set of 97000 contrastive translation pairs based on the WMT English->German translation task, with errors automatically created with simple rules. We report results for a number of systems, and find that recently introduced character-level NMT systems perform better at transliteration than models with byte-pair encoding (BPE) segmentation, but perform more poorly at morphosyntactic agreement, and translating discontiguous units of meaning. accepted at EACL 2017 (v3: minor fix to table 6 description)