It is well known that AI-based language technology—large language models, machine translation systems, multilingual dictionaries, and corpora—is currently limited to three percent of the world’s most widely spoken, financially and politically backed languages. In response, recent efforts have sought to address the “digital language divide” by extending the reach of large language models to “underserved languages.” We show how some of these efforts tend to produce flawed solutions that adhere to a hard-wired representational preference for certain languages, which we call (...) language modeling bias. Language modeling bias is a specific and under-studied form of linguistic bias were language technology by design favors certain languages, dialects, or sociolects with respect to others. We show that language modeling bias can result in systems that, while being precise regarding languages and cultures of dominant powers, are limited in the expression of socio-culturally relevant notions of other communities. We further argue that at the root of this problem lies a systematic tendency of technology developer communities to apply a simplistic understanding of diversity which does not do justice to the more profound differences that languages, and ultimately the communities that speak them, embody. Drawing on the concept of epistemic injustice, we point to the broader ethico-political implications and show how it can lead not only to a disregard for valuable aspects of diversity but also to an under-representation of the needs of marginalized language communities. Finally, we present an alternative socio-technical approach that is designed to tackle some of the analyzed problems. (shrink)

ABSTRACT This paper follows on previous papers which present and evaluate various decision procedures for modal logics. We consider new test sets and systems that have been recently proposed in the literature. This new experimental analysis confirm previous experimental results in showing that SAT based decision procedures, i.e., the procedures built on top of decision procedures for propositional satisfiability, are more efficient than tableau based decision procedures. They also confirm previous evidence of an easy-hard-easy pattern in the satisfiability curve for (...) modal K. Finally, on these tests, SAT based decision procedures are also more efficient than the recently proposed decision procedures based on translation methods. These results contradict some of the claims presented in previous papers by other authors. (shrink)

In the last decade the concept of context has been extensivelyexploited in many research areas, e.g., distributed artificialintelligence, multi agent systems, distributed databases, informationintegration, cognitive science, and epistemology. Three alternative approaches to the formalization of the notion ofcontext have been proposed: Giunchiglia and Serafini's Multi LanguageSystems (ML systems), McCarthy's modal logics of contexts, andGabbay's Labelled Deductive Systems.Previous papers have argued in favor of ML systems with respect to theother approaches. Our aim in this paper is to support these arguments froma (...) theoretical perspective. We provide a very general definition of ML systems, which covers allthe ML systems used in the literature, and we develop a proof theoryfor an important subclass of them: the MR systems. We prove variousimportant results; among other things, we prove a normal form theorem,the sub-formula property, and the decidability of an importantinstance of the class of the MR systems. The paper concludes with a detailed comparison among the alternativeapproaches. (shrink)