Search

The disappointments of the online economy – for instance, user surveillance and systemic labor abuses – stem at least in part from its failures to meaningfully share ownership and governance with relevant stakeholders. Under the banner of ‘platform cooperativism’, an emerging network of cooperative developers, entrepreneurs, labor organizers and scholars is developing an economic ecosystem that seeks to align the ownership and governance of enterprises with the people whose lives are most affected by them. This represents a radical critique of the existing online economy, but it is also a field of experimentation for alternative forms of ownership design. This essay presents and analyzes some of the ways platform cooperativism has begun to generate ownership designs that could serve the platform economy of the future differently than the investor-owned structures that currently prevail.

Compared to the bourgeoning discussions about heresiology in antiquity, especially connected to broader ones about constructions of orthodoxy, very much less has been said about more contemporary examples of this tradition, which, assuredly, is alive and well. Particularly in evangelical Christian circles, there are numerous books available that list, describe, and evaluate the ever-expanding choices people have in the marketplace of religious ideas. Meanwhile, scholars of antiquity have come to realize the significance of this literary genre in the “production” of difference among communities as well as, in certain circumstances, of what came to be called orthodoxies.

This article unravels opportunities that arise from regarding corporate capital as media. First, it reviews relevant overlaps between media theory and legal theory. Second, it considers cases in which capital ownership can be best understood as mediated, affective claims. Third, it explores how recognizing capital as media can open doors to constructive, creative entrepreneurship and activism – to the point of reinventing or undermining received practices of capital ownership altogether. When we think of capital ownership as mediation, we discover our institutional surroundings to be more vulnerable to critical research and more pliable to tactical intervention.

This essay introduces what promise a novel broad-based capital strategy—trusts serving platform users—might hold for the online economy, especially as an enabler of more widespread, organized, and democratic user accountability. It draws on lessons from the experience of employee ownership alongside emerging opportunities for other kinds of broad-based ownership structures. User-oriented trusts could enable meaningful co-governance and profit sharing among essential stakeholders, a prospect that merits research and experimentation.

Hindi grapheme-to-phoneme (G2P) conversion is mostly trivial, with one exception: whether a schwa represented in the orthography is pronounced or unpronounced (deleted). Previous work has attempted to predict schwa deletion in a rule-based fashion using prosodic or phonetic analysis. We present the first statistical schwa deletion classifier for Hindi, which relies solely on the orthography as the input and outperforms previous approaches. We trained our model on a newly-compiled pronunciation lexicon extracted from various online dictionaries. Our best Hindi model achieves state of the art performance, and also achieves good performance on a closely related language, Punjabi, without modification., Comment: 4 pages, 1 figure. To be published in the 2020 Annual Conference of the Association for Computational Linguistics (https://acl2020.org/)

Most current state-of-the art systems for generating English text from Abstract Meaning Representation (AMR) have been evaluated only using automated metrics, such as BLEU, which are known to be problematic for natural language generation. In this work, we present the results of a new human evaluation which collects fluency and adequacy scores, as well as categorization of error types, for several recent AMR generation systems. We discuss the relative quality of these systems and how our results compare to those of automatic metrics, finding that while the metrics are mostly successful in ranking systems overall, collecting human judgments allows for more nuanced comparisons. We also analyze common errors made by these systems., Comment: COLING 2020

Human speakers have an extensive toolkit of ways to express themselves. In this paper, we engage with an idea largely absent from discussions of meaning in natural language understanding--namely, that the way something is expressed reflects different ways of conceptualizing or construing the information being conveyed. We first define this phenomenon more precisely, drawing on considerable prior work in theoretical cognitive semantics and psycholinguistics. We then survey some dimensions of construed meaning and show how insights from construal could inform theoretical and practical work in NLP., Comment: ACL 2020 camera-ready

In the classical view of economic choices, subjects make rational decisions evaluating the costs and benefits of options in order to maximize their overall income. Nonetheless, subjects often fail to reach optimal outcomes. The overt value of an option drives the direction of decisions, but covert factors such as emotion and sunk cost are thought to drive the observed deviations from optimality. Many questions remain to be answered as to 1) which contexts contribute the most to deviation from an optimal solution; and 2) the extent of these effects. In order to tackle these questions, we devised a decision-making task for mice, in which cost and benefit parameters could be independently and flexibly adjusted and for which a tractable optimal solution was known. Comparing mouse behavior with this optimal solution across parameter settings revealed that the factor most strongly contributing to suboptimality was the cost parameter. The quantification of sunk cost, a covert factor implicated in our task design, revealed it as another contributor to suboptimality. In one condition where the large reward option was particularly unattractive and the small reward cost was low, the sunk cost effect and the cost-led suboptimality almost vanished. In this regime and this regime only, mice could be viewed as close to rational. Taken together, our findings support a model whereby parallel neural circuits independently activate and modulate multiple valuation algorithms, and suggest that “rationality” is a task-specific construct even in mice.Significant StatementIrrational factors in economic decision-making often cause significant deviation from optimal outcomes. By devising a flexible economic choice behavior for mice and comparing their behavior with an optimal solution, we investigated overt and covert factors that contributed to suboptimal outcomes and quantified the deviation from optimality. This investigation identified regimes where mice could be viewed as rational or irrational depending upon the parameters in the same task. These findings may provide a platform to investigate biological substrates underlying rational and irrational decision factors.

Decentralization is a term widely used in a variety of contexts, particularly in political science and discourses surrounding the Internet. It is popular today among advocates of blockchain technology. While frequently employed as if it were a technical term, decentralization more reliably appears to operate as a rhetorical strategy that directs attention toward some aspects of a proposed social order and away from others. It is called for far more than it is theorized or consistently defined. This non-specificity has served to draw diverse participants into common political and technological projects. Yet even the most apparently decentralized systems have shown the capacity to produce economically and structurally centralized outcomes. The rhetoric of decentralization thus obscures other aspects of the re-ordering it claims to describe. It steers attention from where concentrations of power are operating, deferring worthwhile debate about how such power should operate. For decentralization to be a reliable concept in formulating future social arrangements and related technologies, it should come with high standards of specificity. It also cannot substitute for anticipating centralization with appropriate mechanisms of accountability.

Research on adpositions and possessives in multiple languages has led to a small inventory of general-purpose meaning classes that disambiguate tokens. Importantly, that work has argued for a principled separation of the semantic role in a scene from the function coded by morphosyntax. Here, we ask whether this approach can be generalized beyond adpositions and possessives to cover all scene participants—including subjects and objects—directly, without reference to a frame lexicon. We present new guidelines for English and the results of an interannotator agreement study.

We propose a coreference annotation scheme as a layer on top of the Universal Conceptual Cognitive Annotation foundational layer, treating units in predicate-argument structure as a basis for entity and event mentions. We argue that this allows coreference annotators to sidestep some of the challenges faced in other schemes, which do not enforce consistency with predicate-argument structure and vary widely in what kinds of mentions they annotate and how. The proposed approach is examined with a pilot annotation study and compared with annotations from other schemes., Comment: Accepted to The First International Workshop on Designing Meaning Representations (DMR), 2019 (in conjunction with ACL 2019)

From China to North Africa to Europe, the earliest institutions of higher education arose as a set of strategies to protect and encourage practices of productive leisure. These were knowledge-production practices the surrounding social and economic order might not otherwise have seen fit to value: self-directed time organized to meet collectively agreed-upon standards. Practices of productive leisure have also become targets for colonial regimes seeking to undermine local civilizations for their own ends. Today, in an era of growing labor precarity and the specter of technological unemployment, higher education is poised to ensure that the benefits of the transformation underway circulate to a broader set of people than is currently the case. Updating the medieval guild structure of universities with cooperative enterprise, and platform cooperativism in particular, can facilitate the capture and sharing of economic surplus. By protecting productive leisure through higher education, automation can yield widespread dividends of liberated time.

From China to North Africa to Europe, the earliest institutions of higher education arose as a set of strategies to protect and encourage practices of productive leisure. These were knowledge-production practices the surrounding social and economic order might not otherwise have seen fit to value: self-directed time organized to meet collectively agreed-upon standards. Practices of productive leisure have also become targets for colonial regimes seeking to undermine local civilizations for their own ends. Today, in an era of growing labor precarity and the specter of technological unemployment, higher education is poised to ensure that the benefits of the transformation underway circulate to a broader set of people than is currently the case. Updating the medieval guild structure of universities with cooperative enterprise, and platform cooperativism in particular, can facilitate the capture and sharing of economic surplus. By protecting productive leisure through higher education, automation can yield widespread dividends of liberated time.

T he genre of testimony is one that a serious writer would do best to avoid. By testimony I mean a very particular subset of first-person narrative—this isn’t autobiography (of one’s life as such), or reportage (of one’s adventures), or confession (of one’s misadventures), or testimonial (of one’s experiences with certain products). Each of those genres can be adapted, free of cringing, to a modern, literary, secular voice. One can write in them, post the result on the Internet, then afterward live to write another genre. Testimony is different. That’s because writers are supposed to be in control, aware of themselves and those around them. Supremely so— that’s why we’re the writers. But testimony doesn’t permit this. Whether one testifies from the stand in a courtroom or the sanctuary of a church, testimony is about forces beyond one’s control or comprehension. In court, testimony takes place in the midst of an imposing and perplexing and grinding process by which justice is purported to be done, and over which the witness has as little control as the lawyer class can make possible. Such a testimony can be only part of the story, and it will be heard only through a mysterious combination of performative and bureaucratic filters. Testimony coming from a Christian has similar features. This kind of testimony is a story about what God has done in one’s life; the main character in such a story is supposed to be God, but since the testifier is a human being and not God, God’s activities present themselves only through what the testifier claims God has done in her or his life. A writer conveying such experience has little opportunity to take credit for

Theories of discourse coherence posit relations between discourse segments as a key feature of coherent text. Our prior work suggests that multiple discourse relations can be simultaneously operative between two segments for reasons not predicted by the literature. Here we test how this joint presence can lead participants to endorse seemingly divergent conjunctions (e.g., BUT and SO) to express the link they see between two segments. These apparent divergences are not symptomatic of participant naivety or bias, but arise reliably from the concurrent availability of multiple relations between segments – some available through explicit signals and some via inference. We believe that these new results can both inform future progress in theoretical work on discourse coherence and lead to higher levels of performance in discourse parsing.

After decades of neglect, antitrust is once again a topic of public debate. Proponents of reviving antitrust have called for abandoning the narrow consumer welfare model and embracing a broader set of objectives. One essential element that has been overlooked thus far is the ownership structure of the firm itself. The dominant model of investor-owned business and associated philosophy of shareholder wealth maximization exacerbate the pernicious effects of market power. In contrast, cooperative ownership models can mitigate the effects of monopoly and oligopoly and advance the interests of consumers, workers, small business owners, and citizens. The promotion of competition among large firms should be paired with support for democratic cooperation within firms. Antitrust law has had a complicated history and relationship with cooperative enterprise. Corporations threatened by cooperatives have used the antitrust laws to frustrate the growth of these alternative businesses. To insulate cooperatives from the antitrust threat, Congress has enacted exemptions to protect cooperative entities, notably a general immunity for farm cooperatives in the 1922 Capper-Volstead Act. As part of an agenda to tame corporate monopoly, all three branches of the federal government and the states should revisit these ideas and seek to protect and enable the cooperative model across the economy. While protections that farmers fought for a century ago may seem obsolete in an era of big-box retail and online platforms, matters of ownership design have at least as much relevance today and should be a part of the antimonopoly arsenal.

A new Python API, integrated within the NLTK suite, offers access to the FrameNet 1.7 lexical database. The lexicon (structured in terms of frames) as well as annotated sentences can be processed programatically, or browsed with human-readable displays via the interactive Python prompt., Comment: EMNLP 2017 Demo

We consider the semantics of prepositions, revisiting a broad-coverage annotation scheme used for annotating all 4,250 preposition tokens in a 55,000 word corpus of English. Attempts to apply the scheme to adpositions and case markers in other languages, as well as some problematic cases in English, have led us to reconsider the assumption that an adposition’s lexical contribution is equivalent to the role/relation that it mediates. Our proposal is to embrace the potential for construal in adposition use, expressing such phenomena directly at the token level to manage complexity and avoid sense proliferation. We suggest a framework to represent both the scene role and the adposition’s lexical function so they can be annotated at scale—supporting automatic, statistical processing of domain-general language—and discuss how this representation would allow for a simpler inventory of labels.

The goal of this research is to add to the knowledge base of bamboo-reinforced concrete (BRC) by focusing on a unique stirrup design and testing the lap-splicing of flexural reinforcement. Preliminary tests of the mechanical properties of bamboo were performed, and the results have been used in the design and construction of bamboo cages to reinforce concrete beams. This paper presents the preliminary results of tensile tests and pull-out tests on Moso (Phyllostachys edulis) bamboo. The test results show that bamboo is a viable alternative to steel as tensile reinforcement for concrete structures. As part of the on-going research, 14 Moso bamboo (Phyllostachys edulis) reinforced concrete beams will be cast and tested under monotonic gravity loading in early 2014 to assess the performance of bamboo as flexural reinforcement, shear reinforcement and lap-splicing.

We describe a new dependency parser for English tweets, TWEEBOPARSER. The parser builds on several contributions: new syntactic annotations for a corpus of tweets (TWEEBANK), with conventions informed by the domain; adaptations to a statistical parsing algorithm; and a new approach to exploiting out-of-domain Penn Treebank data. Our experiments show that the parser achieves over 80% unlabeled attachment accuracy on our new, high-quality test set and measure the benefit of our contributions. Our dataset and parser can be found at http://www.ark.cs.cmu.edu/TweetNLP.