David Strohmaier

No natural language has contrafactive attitude verbs. Because factives are universal across natural languages, this means that there is a major asymmetry between contrafactives and factives. We previously hypothesised that this asymmetry arises partly because the meaning of contrafactives is significantly harder to learn than that of factives. Here we test this hypothesis by using a production-oriented computational experiment that overcomes two limitations of our previous experiments. We find that our results do not support our previous hypothesis.

For most of its history, decision theory has investigated the rational choices of humans under the assumption of static preferences. Human preferences, however, change. In recent years, decision theory has increasingly acknowledged the reality of preference change throughout life. This Element provides an accessible introduction and new contributions to the debates on preference change. It is divided into three chapters. In the first chapter, the authors discuss what preference change is and whether we can integrate it into decision theory. In the second chapter, they present models of preference change, including a novel proposal of their own. In the third and final chapter, they discuss how we can rationally choose a course of action when our preferences might change. Both the transformative experience literature and recent work on choosing for changing selves are discussed.

Estimating word complexity is a well-established task in computer-assisted language learning. So far, however, complexity estimation has been largely limited to comprehension. This neglects words that are easy to comprehend, but hard to produce. We introduce semantic error prediction (SEP) as a novel task that assesses the production complexity of content words. Given the corrected version of a learner-produced text, a system has to predict which content words replace tokens from the original text. We present and analyse one example of such a semantic error prediction dataset, which we generate from an error correction dataset. As neural baselines, we use BERT, RoBERTa, and LLAMA2 embeddings for SEP. We show that our models can already improve downstream applications, such as predicting essay vocabulary scores.

The Sense Complexity Dataset (SeCoDa) provides a corpus that is annotated jointly for complexity and word senses. It thus provides a valuable resource for both word sense disambiguation and the task of complex word identification. The intention is that this dataset will be used to identify complexity at the level of word senses rather than word tokens. For word sense annotation SeCoDa uses a hierarchical scheme that is based on information available in the Cambridge Advanced Learner’s Dictionary. This way we can offer more coarse-grained senses than directly available in WordNet.

Richard Holton has drawn attention to a new semantic universal, according to which (almost) no natural language has contrafactive attitude verbs. This semantic universal is part of an asymmetry between factive and contrafactive attitude verbs. Whilst factives are abundant, contrafactives are scarce. We propose that this asymmetry is partly due to a difference in learnability. The meaning of contrafactives is significantly harder to learn than that of factives. We tested our hypothesis by conducting a computational experiment using an artificial neural network. The results of this experiment support our hypothesis.

Holton has drawn attention to a new semantic universal, according to which no natural language has contrafactive attitude verbs. Because factives are universal across natural languages, Holton’s universal is part of a major asymmetry between factive and contrafactive attitude verbs. We previously proposed that this asymmetry arises partly because the meaning of contrafactives is significantly harder to learn than that of factives. Here we extend our work by describing an additional computational experiment that further supports our hypothesis.

Organisations are computing systems. The university’s sports centre is a computing system for managing sports teams and facilities. The tenure committee is a computing system for assigning tenure status. Despite an increasing number of publications in group ontology, the computational nature of organisations has not been recognised. The present paper is the first in this debate to propose a theory of organisations as groups structured for computing. I begin by describing the current situation in group ontology and by spelling out the thesis in more detail. I then present the example of a sports centre to illustrate why one might intuitively think of organisations as computing systems. To substantiate the thesis, I introduce Piccinini’s restrictive analysis of physical computation. As I show, organisations meet all criteria for being computing systems. Organisations are structured groups with the function of manipulating medium-independent vehicles according to rules. Furthermore, I argue for the modal claim that this is a necessary feature of organisations. Having sketched the computational account of organisations, I compare it to other proposals in the literature.

The ontology of social objects and facts remains a field of continued controversy. This situation complicates the life of social scientists who seek to make predictive models of social phenomena. For the purposes of modelling a social phenomenon, we would like to avoid having to make any controversial ontological commitments. The overwhelming majority of models in the social sciences, including statistical models, are built upon ontological assumptions that can be questioned. Recently, however, artificial neural networks have made their way into the social sciences, raising the question whether they can avoid controversial ontological assumptions. ANNs are largely distinguished from other statistical and machine learning techniques by being a representation-learning technique. That is, researchers can let the neural networks select which features of the data to use for internal representation instead of imposing their preconceptions. On this basis, I argue that neural networks can avoid ontological assumptions to a greater degree than common statistical models in the social sciences. I then go on, however, to establish that ANNs are not ontologically innocent either. The use of ANNs in the social sciences introduces ontological assumptions typically in at least two ways, via the input and via the architecture.

Social kinds are heterogeneous. As a consequence of this diversity, some authors have sought to identify and analyse different kinds of social kinds. One distinct kind of social kinds, however, has not yet received sufficient attention. I propose that there exists a class of social-computation-supporting kinds, or SCS-kinds for short. These SCS-kinds are united by the function of enabling computations implemented by social groups. Examples of such SCS-kinds arereimbursement form,US dollar bill,chair of the board. I will analyse SCS-kinds, contrast my analysis with theories of institutional kinds, and discuss the benefits of investigating SCS-kinds.

Despite having faced severe criticism in the past, mereological approaches to group ontology, which argue that groups are wholes and that groups members are parts, have recently managed a comeback. Authors such as Katherine Ritchie and Paul Sheehy have applied neo-Aristotelian mereology to groups, and Katherine Hawley has defended mereological approaches against the standard objections in the literature. The present paper develops the mereological approaches to group ontology further and proposes an analysis of group membership as parthood plus further restrictions. While all mereological accounts agree that group members are parts of the group, it has become clear that this analysis is insufficient. I discuss three proposals to develop the mereological analysis of group membership and then put forward a combined solution to the puzzle. According to my proposal, the members of a reading group are agents who are part of the group and have been designated to contribute to the group.

Two theories dominate the current debate on group agency: functionalism, as endorsed by Bryce Huebner and Brian Epstein, and interpretivism, as defended by Deborah Tollefsen, and Christian List and Philip Pettit. In this paper, I will give a new argument to favour functionalism over interpretivism. I discuss a class of cases which the former, but not the latter, can accommodate. Two features characterise this class: First, distinct groups coincide, that is numerically distinct groups share all their members at all time. Second, we have access to the inner mechanisms of the groups agents, because members know what they have decided on. I construct a counterexample with these features allowing me to reject interpretivism about group agency in favour of functionalism.