As natural language processing tools powered by big data become increasingly ubiquitous, questions of how to design, develop, and manage these tools and their impacts on diverse populations are pressing. We propose utilizing the concept of linguistic justice—the realization of equitable access to social and political life regardless of language—to provide a framework for examining natural language processing tools that learn from and use human language data. To support linguistic justice, we argue that natural language processing tools must be examined (...) not only from the perspective of a privileged, majority language user, but also from the perspectives of minoritized language users. Considering such perspectives can help to surface areas in which the data used within natural language processing tools may be working against linguistic justice by failing to provide access to information, services, or opportunities in users’ language of choice, underperforming for certain linguistic groups, or advancing harmful stereotypes that can lead to negative life outcomes for members of marginalized groups. At the same time, this framework can help to illuminate ways that these shortcomings can be addressed and allow us to use inclusive language data and approaches to leverage natural language processing technologies that advance linguistic justice. (shrink)

As modern deep networks become more complex, and get closer to human‐like capabilities in certain domains, the question arises as to how the representations and decision rules they learn compare to the ones in humans. In this work, we study representations of sentences in one such artificial system for natural language processing. We first present a diagnostic test dataset to examine the degree of abstract composable structure represented. Analyzing performance on these diagnostic tests indicates a lack of systematicity in representations (...) and decision rules, and reveals a set of heuristic strategies. We then investigate the effect of training distribution on learning these heuristic strategies, and we study changes in these representations with various augmentations to the training set. Our results reveal parallels to the analogous representations in people. We find that these systems can learn abstract rules and generalize them to new contexts under certain circumstances—similar to human zero‐shot reasoning. However, we also note some shortcomings in this generalization behavior—similar to human judgment errors like belief bias. Studying these parallels suggests new ways to understand psychological phenomena in humans as well as informs best strategies for building artificial intelligence with human‐like language understanding. (shrink)

Psychologists and neuroscientists extensively rely on computational models for studying and analyzing the human mind. Traditionally, such computational models have been hand-designed by expert researchers. Two prominent examples are cognitive architectures and Bayesian models of cognition. While the former requires the specification of a fixed set of computational structures and a definition of how these structures interact with each other, the latter necessitates the commitment to a particular prior and a likelihood function which – in combination with Bayes’ rule – (...) determine the model's behavior. In recent years, a new framework has established itself as a promising tool for building models of human cognition: the framework of meta-learning. In contrast to the previously mentioned model classes, meta-learned models acquire their inductive biases from experience, i.e., by repeatedly interacting with an environment. However, a coherent research program around meta-learned models of cognition is still missing to this day. The purpose of this article is to synthesize previous work in this field and establish such a research program. We accomplish this by pointing out that meta-learning can be used to construct Bayes-optimal learning algorithms, allowing us to draw strong connections to the rational analysis of cognition. We then discuss several advantages of the meta-learning framework over traditional methods and reexamine prior work in the context of these new insights. (shrink)

C’est la CEPT (a.k.a. “Emptiness within Emptiness”) as open-ended, performance-based cinematic project grounded in ambient architectural and scenographic utility, utilizes a semi-abandoned building (badminton court) in Ahmedabad, India, origin of the School of Architecture (c.1962), later CEPT University, for a polemical and tragi-comic investigation of the vagaries of institutional memory, inclusive of intentional repressions. The pseudo-psychoanalytical prospects of the project question whether “emptiness” is a concept relative to subjective versus objective states. By hypothetically placing one form of emptiness within another (...) – i.e., via submission of the theatrical-cinematic work to a design competition for an installation at the EMST, a never-opened museum in Athens, Greece – an empty “ancient” building in India (the East) is conjoined with an empty new (i.e., never-opened) building in Greece (the West), the conflation becoming a means for doubling the subjective states under investigation and inducing hoped-for catharsis on all accounts. (shrink)

In this volume well-known scholars from India and Latin America - Enrique Dussel, Madhu Dubey, Walter D. Mignolo, and Sudipta Sen, to name a few - discuss the concepts of modernity and colonialism and describe how the two relate to each other. This second edition to the volume comes with a new introduction which extends and critically supplements the discussion in the earlier introduction to the volume. It explores the vital impact of the colonial pasts of India, Mexico, China, and (...) even the Unites States, on the processes through which these countries have become modern. The collection is unique, as it brings together a range of disciplines and perspectives. The topics discussed include the Zapatista movement in Southern Mexico, the image of the South in recent African-American literature, the theories of Andre Gunder Frank about the early modernization of Asian countries, and the contradictions of the colonial state in India. (shrink)