Online Speech Research Articles

The Impact of Culture and Gender on the Production of Online Speech Acts among Jordanian Facebook Users

This study of speech acts offers deep insights into the social structure and gender differences of any speech community. Most relevant research on online speech acts has shed much light on Western speech communities, neglecting the speech act behavior of Arabic speaking communities. This study aimed to examine the influence of gender differences and the Jordanian cultural norms on the use of speech acts in Facebook Status Updates (FSUs). A total of 1718 FSUs were collected over a period of 2 months. Then, they were categorized according to Searle’s speech acts taxonomy. Results showed that women made an average of 6 updates to their Facebook status, while men averaged four. In addition, the results revealed that females and males differ in their linguistic repertoires. Male participants were inclined to use more assertive speech acts, whereas females were more likely to use expressive speech acts. Islam and tribalism are factors that were found to play an important role in defining the cultural norms of the Jordanian speech community. .

Individual differences in subphonemic sensitivity and phonological skills.

Many studies have established a link between phonological abilities (indexed by phonological awareness and phonological memory tasks) and typical and atypical reading development. Individuals who perform poorly on phonological assessments have been mostly assumed to have underspecified (or "fuzzy") phonological representations, with typical phonemic categories, but with greater category overlap due to imprecise encoding. An alternative posits that poor readers have overspecified phonological representations, with speech sounds perceived allophonically (phonetically distinct variants of a single phonemic category). On both accounts, mismatch between phonological categories and orthography leads to reading difficulty. Here, we consider the implications of these accounts for online speech processing. We used eye tracking and an individual differences approach to assess sensitivity to subphonemic detail in a community sample of young adults with a wide range of reading-related skills. Subphonemic sensitivity inversely correlated with meta-phonological task performance, consistent with overspecification.

Platforms, the First Amendment and Online Speech: Regulating the Filters

In recent years, online platforms have given rise to multiple discussions about what their role is, what their role should be, and whether they should be regulated. The complex nature of these private entities makes it very challenging to place them in a single descriptive category with existing rules. In today’s information environment, social media platforms have become a platform press by providing hosting as well as navigation and delivery of public expression, much of which is done through machine learning algorithms. This article argues that there is a subset of algorithms that social media platforms use to filter public expression, which can be regulated without constitutional objections. A distinction is drawn between algorithms that curate speech for hosting purposes and those that curate for navigation purposes, and it is argued that content navigation algorithms, because of their function, deserve separate constitutional treatment. By analyzing the platforms’ functions independently from one another, this paper constructs a doctrinal and normative framework that can be used to navigate some of the complexity.

Regulating Habit-Forming Technology

Tech developers, like slot machine designers, strive to maximize the user’s “time on device.” They do so by designing habit-forming products – products that draw consciously on the same behavioral design strategies that the casino industry pioneered. The predictable result is that most tech users spend more “time on device” than they would like and a substantial minority develop life-changing behavioral problems similar to problem gambling. Other countries have begun to regulate habit-forming tech, and American jurisdictions may soon follow suit. Several state legislatures today are considering bills to regulate “loot boxes,” a highly addictive slot-machine-like mechanic that is common in online video games. The Federal Trade Commission has also announced an investigation into the practice. As public concern mounts, it is surprisingly easy to envision consumer regulation extending beyond video games to other types of apps. Just as tobacco regulations might prohibit brightly-colored packaging and fruity flavors, a social media regulation might limit the use of red notification badges or “streaks” that reward users for daily use. It is unclear how much of this regulation could survive First Amendment scrutiny; software, unlike other consumer products, is widely understood as a form of protected “expression.” But it is also unclear that well-drawn laws to combat compulsive technology use would seriously threaten First Amendment values. At a very low cost to the expressive interests of the tech companies, these laws may well enhance the quality and efficacy of online speech by mitigating distraction and promoting deliberation.

Lexical processing depends on sublexical processing: Evidence from the visual world paradigm and aphasia.

Some early studies of people with aphasia reported strikingly better performance on lexical than on sublexical speech perception tasks. These findings challenged the claim that lexical processing depends on sublexical processing and suggested that acoustic information could be mapped directly to lexical representations. However, Dial and Martin (Neuropsychologia 96: 192-212, 2017) argued that these studies failed to match the discriminability of targets and distractors for the sublexical and lexical stimuli and showed that when using closely matched tasks with natural speech tokens, no patient performed substantially better at the lexical than at the sublexical processing task. In the current study, we sought to provide converging evidence for the dependence of lexical on sublexical processing by examining the perception of synthetic speech stimuli varied on a voice-onset time continuum using eye-tracking methodology, which is sensitive to online speech perception processes. Eight individuals with aphasia and ten age-matched controls completed two visual world paradigm tasks: phoneme (sublexical) and word (lexical) identification. For both identification and eye-movement data, strong correlations were observed between the sublexical and lexical tasks. Critically, no patient within the control range on the lexical task was impaired on the sublexical task. Overall, the current study supports the claim that lexical processing depends on sublexical processing. Implications for inferring deficits in people with aphasia and the use of sublexical tasks to assess sublexical processing are also discussed.

Facebook v. Sullivan: Building Constitutional Law for Online Speech

In the United States, there are now two systems to adjudicate disputes about harmful speech. The first is older and more established: the legal system in which judges apply constitutional law to limit tort claims alleging injuries caused by speech. The second is newer and less familiar: the content-moderation system in which platforms like Facebook implement the rules that govern online speech. These platforms are not bound by the First Amendment. But, as it turns out, they rely on many of the tools used by courts to resolve tensions between regulating harmful speech and preserving free expression—particularly the entangled concepts of “public figures” and “newsworthiness.” This Article offers the first empirical analysis of how judges and content moderators have used these two concepts to shape the boundaries of free speech. It first introduces the legal doctrines developed by the “Old Governors,” exploring how courts have shaped the constitutional concepts of public figures and newsworthiness in the face of tort claims for defamation, invasion of privacy, and intentional infliction of emotional distress. The Article then turns to the “New Governors” and examines how Facebook’s content-moderation system channeled elements of the courts’ reasoning for imposing First Amendment limits on tort liability. By exposing the similarities and differences between how the two systems have understood these concepts, this Article offers lessons for both courts and platforms as they confront new challenges posed by online speech. It exposes the pitfalls of using algorithms to identify public figures; explores the diminished utility of setting rules based on voluntary involvement in public debate; and analyzes the dangers of ad hoc and unaccountable newsworthiness determinations. Both courts and platforms must adapt to the new speech ecosystem that companies like Facebook have helped create, particularly the way that viral content has shifted normative intuitions about who deserves harsher rules in disputes about harmful speech, be it in law or content moderation. Finally, the Article concludes by exploring what this comparison reveals about the structural role platforms play in today’s speech ecosystem and how it illuminates new solutions. These platforms act as legislature, executive, judiciary, and press—but without any separation of powers to establish checks and balances. A change to this model is already occurring at one platform: Facebook is creating a new Oversight Board that will hopefully provide due process to users on the platform’s speech decisions and transparency about how content-moderation policy is made, including how concepts related to newsworthiness and public figures are applied.

Speech Across Borders

As both governments and tech companies seek to regulate speech online, these efforts raise critical, and contested, questions about how far those regulations can and should extend. Is it enough to take down or delink material in a geographically segmented way? Or can and should tech companies be ordered to takedown or delink unsavory content across their entire platforms—no matter who is posting the material or where the unwanted content is viewed? How do we deal with conflicting speech norms across borders? And how do we protect against the most censor-prone nation effectively setting global speech rules? These questions were recently addressed in two high-profile judgments from the European Court of Justice and were the subject of ongoing litigation that pitted Canadian and U.S. courts against one another. Meanwhile, a new form of geographically-segmented speech regulation is emerging—pursuant to which speech is limited based on who is speaking and from where, as opposed to what is being said. This Article examines the ways in which norms regarding speech, privacy, and a range of other rights conflict across borders, the power of private sector players in adjudicating and resolving these conflicts, the ways in which governments are seeking to harness this power on a global scale, and the broader implications for individual rights. It offers a nuanced approach that identifies the multiple competing interests at stake—recognizing both the ways in which global takedowns and delistings can, at times, be a critical means of protecting key interests, and the risk of over-censorship and forced uniformity that can result.

Cyber Safe Curricula and Online Harassment

AbstractIn this article, Heather Greenhalgh‐Spencer analyzes three examples of curricula aimed at cultivating good cyber behavior and countering online harassment. These curricula are branded as addressing civic needs, cultivating civic duty, and developing skills to help students understand and react to online threats. Here, Greenhalgh‐Spencer offers two critiques of these, and other similar, cyber safe curricula. First, current cyber safe curricula do not adequately address the ways that identity — particularly gender and sexuality — shape one's experience of online spaces and exposure to online threats. Second, current cyber safe curricula do not adequately consider nuances of how to “react” to online threats. The preponderance of cyber curricula teach students that all forms of offensive speech online should be abolished or pushed against. These curricula miss the opportunity to cultivate conversations around the tensions between the need to feel safe and the need to protect freedoms of speech.

A gestão do discurso de ódio nas plataformas de redes sociais digitais: um comparativo entre Facebook, Twitter e Youtube

O impasse social sobre discurso de ódio e liberdade de expressão na Internet tem impulsionado os Sites de Redes Sociais (SRS) a intensificarem suas políticas de moderação de conteúdo. A gestão do conteúdo de ódio em plataformas como Facebook, Twitter e Youtube é tão complexa, haja vista seu caráter multifacetado e o grande número de interagentes, que executivos destas empresas assumem a ineficiência de seus recursos (humanos e tecnológicos) na tentativa de controlar o escalonamento, duração, difusão e circunspecção de crimes e discursos odientos. O problema tem despertado a atenção de governos e organizações civis, que por sua vez aumentam a pressão sobre as plataformas no intuito de melhorarem suas escolhas editoriais e sua logística de monitoramento e remoção deste tipo de interação. Neste contexto, este trabalho tem como objetivo comparar as ações realizadas pelo Facebook, Twitter e Youtube no que tange a formulação e ampliação de políticas e decisões sobre conteúdo individual de ódio. Para isto, foi realizado uma análise histórica (2015-2018) de dados secundários das políticas e termos de comunidade específicos de cada SRS numa análise das tomadas de decisão com os cinco tópicos do termo de compromisso que as empresas assumiram com a ADL -Liga Anti-Difamação, em 2013, no combate ao discurso de ódio online. Os resultados apontam o Facebook como o SRS que mais investiu em estratégias de combate a intolerância e incivilidade online, apesar da empresa não deixar claro os métodos empregados para tal fim. De modo geral, todas as plataformas evoluíram na estrutura operacional de denúncia de conteúdo odiento, mas se mostraram ineficientes em moderação, remoção e contenção da propagação do cyberhate.

The German Act on Improving Law Enforcement on Social Networks (NetzDG): A Blunt Sword?

On 1 October 2017, the German law with the rather cumbersome title Netzwerkdurchsetzungsgesetz, which literally translates to “Network Enforcement Act”, entered into force. This federal law aims to improve law enforcement in social networks by calling providers of social net-works to account in order to act on online speech that is punishable under domestic criminal law. During the legislative process and in the materials accompanying the law, the legislator repeatedly stressed that the law is necessary to respond to a (perceived) proliferation of hate speech and fake news on social media. Facebook, one of the very few social networks to which the Act will apply, expressed its concerns harshly during the legislative proceedings: the bill was deemed imprecise, unconstitutional and a threat for freedom of opinion in Germany. The bill did not only stir criticism from social media companies, also many legal scholars and the scientific research service of the German Parliament raised their concerns. Against the severe criticism, the Act also attracted attention beyond Germany with in particular the United Kingdom and France eyeing on its impact in practice. With the first anniversary of the NetzDG, this paper will look at the past, present and future of the highly controversial law and examine the impact of the law that has been rated as to seriously impair human rights online when it entered into force.

Differences in Working Memory Capacity Affect Online Spoken Word Recognition: Evidence From Eye Movements.

Individual differences in working memory capacity have been gaining recognition as playing an important role in speech comprehension, especially in noisy environments. Using the visual world eye-tracking paradigm, a recent study by Hadar and coworkers found that online spoken word recognition was slowed when listeners were required to retain in memory a list of four spoken digits (high load) compared with only one (low load). In the current study, we recognized that the influence of a digit preload might be greater for individuals who have a more limited memory span. We compared participants with higher and lower memory spans on the time course for spoken word recognition by testing eye-fixations on a named object, relative to fixations on an object whose name shared phonology with the named object. Results show that when a low load was imposed, differences in memory span had no effect on the time course of preferential fixations. However, with a high load, listeners with lower span were delayed by ∼550 ms in discriminating target from sound-sharing competitors, relative to higher span listeners. This follows an assumption that the interference effect of a memory preload is not a fixed value, but rather, its effect is greater for individuals with a smaller memory span. Interestingly, span differences affected the timeline for spoken word recognition in noise, but not offline accuracy. This highlights the significance of using eye-tracking as a measure for online speech processing. Results further emphasize the importance of considering differences in cognitive capacity, even when testing normal hearing young adults.

Public apologies and intercultural communication: Perceptions of Chinese and American students

The primary purpose of this study was to investigate how culture impacts recipients’ perceptions of the effectiveness of a public apology (Netflix apology). Data was collected through a survey instrument that included both Likert Scale items and short response items. A selected number of participants from each group were also interviewed. Quantitative and qualitative analysis of data were conducted. The study also aimed to identify specific verbal and non-verbal cues that correlate with students’ evaluations of this public apology. Overall, the participants evaluated the Netflix apology as ineffective in terms of verbal as well as non-verbal strategies used. Findings show that cultural variations regarding the appropriate non-verbal cues significantly affected the participants’ evaluations of the apology. Chinese emphasized the importance of a formal setting, professional dress, bowing posture and remorseful facial expressions, whereas Americans emphasized the importance of eye contact and body posture embodying attentiveness. The majority of the participants commented that the apologizer should have offered compensation as a means of taking responsibility for the offense. The study has teaching implications for raising intercultural competency among the students and business leaders, particularly regarding the use of online speech events in teaching scenarios.

State regulation of online speech in Russia: the role of internet infrastructure owners

State regulation of online speech in Russia: the role of internet infrastructure owners

(Re)embracing social responsibility theory as a basis for media speech: shifting the normative paradigm for a modern media

Dave Egger’s fictional book The Circle tells the story of an all-powerful new media company of the same name that seeks to totally monopolise its market and remake the world in its image. To achieve this The Circle advocates the unregulated sharing of all information, at all times, regardless of its source and irrespective of the consequences for individuals, society and the state. Although the dystopian view of reality presented by the book is perhaps slightly extreme, it does not take any great leap of faith to see how we could all end up as ‘Circlers’, particularly because the underlying normative rationale that drives The Circle is what currently underpins online speech in reality. Libertarianism and the inherently libertarian arguments from truth and the marketplace of ideas have historically underpinned the notion of the Fourth Estate and have a ‘hold’ on First Amendment jurisprudence. In recent years, libertarianism has emerged as the de facto normative paradigm for internet and social media speech worldwide. Although the theory’s dominant position fits with the perceived ethos of social media platforms such as Facebook, its philosophical foundations are based on nineteenth and early twentieth-century means of communication. Consequently, as illustrated by issues such as filter bubbles and Facebook’s reaction to fake news (bringing in a third-party fact-checking company) which conflicts with the platform’s libertarian ideology, as well as the European Court of Human Rights consistently placing the argument from democracy at the heart of its Article 10 ECHR jurisprudence, rather than the argument from truth and marketplace of ideas, this normative framework is idealistic as opposed to being realistic. Therefore, it is not suitable for twenty-first-century free speech and the modern media, of which social media is no longer an outlier, but a central component. Thus, this paper advances the argument that a normative and philosophical framework for media speech, based on social responsibility theory and the argument from democratic self-governance, is more suitable for the modern media than libertarianism. Further, it justifies a coercive regulatory regime that also preserves media freedom.

Hate Speech Online: an (Intractable) Contemporary Challenge?

Internet users generate billions of pieces of online content weekly across a number of social media platforms, and that content includes hate speech. How to respond to hate speech online is a question that is troubling democracies all over the world and there is no easy solution in sight. The question of hate speech has long given rise to dispute in international human rights law, a dispute that arises because the protection of freedom of speech on the one hand and the prohibition of hate speech on the other are rooted in different normative principles that need to be accommodated. This enduring dispute makes it particularly difficult to design solutions to the problem of hate speech online. The article describes and assesses the current rules regulating hate speech online in four jurisdictions, the USA, the UK, Europe and Germany and suggests that it is not clear that any of the systems satisfactorily address the issue of hate speech online.

An audio fingerprinting extraction algorithm based on lifting wavelet packet and improved optimal-basis selection

Audio fingerprinting technology is widely applied to the analysis and processing of digital signal, especially in the application of speech recognition which is one of the most popular fields of the intelligent multimedia and artificial intelligence. Traditional audio fingerprinting extraction algorithm is based on the decomposition and reconstruction of the wavelet packet. But the requirement of computational capacity and memory is so large. So this paper proposed an algorithm which is based on the lifting wavelet packet and the improved optimal-basis selection to find the coefficient of optimal wavelet packet. Then the average of the logarithmic energy entropy is adopted as the characteristic parameter. And the capacity of computing and memory is better than the traditional algorithm because of the lifting wavelet packet which is more suitable for processing of speech online and the design of intelligent multimedia. And the experiment results indicate that this algorithm is not only robust for the audio which is handled by some kinds of methods and can reflect the overall characteristics of the audio very well, but also has good distinguishability between different audio.

Privatised enforcement and the right to freedom of expression in a world confronted with terrorism propaganda online

The purpose of this paper is to explore the risks of privatised enforcement in the field of terrorism propaganda, stemming from the EU Code of conduct on countering illegal hate speech online. By shedding light on this Code, the author argues that implementation of it may undermine the rule of law and give rise to private censorship. In order to outweigh these risks, IT companies should improve their transparency, especially towards users whose content have been affected. Where automated means are used, the companies should always have in place some form of human intervention in order to contextualise posts. At the EU level, the Commission should provide IT companies with clearer guidelines regarding their liability exemption under the e-Commerce Directive. This would help prevent a race-to-the bottom where intermediaries choose to interpret and apply the most stringent national laws in order to secure at utmost their liability. The paper further articulates on the fine line that exists between ‘terrorist content’ and ‘illegal hate speech’ and the need for more detailed definitions.

Supporting deafblind in congregational prayer using speech recognition and vibro-tactile stimuli

Abstract Human behavior inference through smart pervasive technologies is a new research avenue. Sophistication in interaction modalities with self-customization based on context makes interaction easier in natural settings. Despite greater sophistication, disabled and elderly person still face difficulties in interaction with smart devices. The interaction generally depends on two or more sensory channels simultaneously, which is a big concern for persons with impaired vision and hearing. Person with dual sensory loss face problem in interaction with smart devices. Interaction can be effectively possible with this population group using alternative tactile sensory channel and vibro-tactile stimuli could be a great alternative for communication in natural settings to accommodate their needs. Feature-rich handheld and wearable smart devices could be the cost-effective substitute for assisting in ADL in natural settings for disabled and elderly person. Religious worships are important part of human life. By exploiting the advanced sensing technologies in Smartphone, an assistive prayer model is developed which is aimed at assisting deafblind person in semi-isolated natural settings, in a cost-effective manner which emphasizes on utilization of alternative sensory channels of a human. The model is validated through an App SmartPrayerAid aimed to assist deafblind Muslim in offering Islamic worship Namaz in congregational setting. Comprehensive evaluation of the App was conducted to prove the validity of the model and the aptness of the smart devices for assisting disabled person in natural settings. The evaluation was conducted with deafblind, non-deafblind and non-disabled subjects in controlled lab environment. Results indicated that speech recognition was achieved between 77.76% in real environment, while up to 96.29% in controlled lab environment using online speech recognition in Arabic. Effectiveness of Vibro-tactile stimuli perception using Smartphone on different body parts was also evaluated and accuracy in detection observed on front pocket of shirt was 84.29%, front pocket of the trousers was 92.14% and on upper arm was 52.43% and on wrist of the right hand using Smartwatch was observed 97.62%. The model can be utilized in other natural settings in similar situations after slight modification to assist a deafblind person in performing ADL and other social activities.

Campaign Finance Disclosure

Campaign finance disclosure is the last (somewhat) robust regulation we have in American campaign finance, and it is under threat. We urgently need more research on disclosure. Regulatory complexity makes studying campaign finance disclosure daunting. It also creates so-called dark money and anonymous speech online. Scholars must understand the existing regulatory loopholes as they plan studies to avoid biased estimates and understand the conditions in which their results generalize to a broader population. The court's disclosure jurisprudence is thin and based on largely unproven assumptions. As the research on campaign finance disclosure matures, scholars should take a broad view of the costs and benefits of disclosure, rather than the narrow, court-led focus many studies have had until now. We must also take seriously the ways in which cognitive limitations can limit the benefits of disclosure. I explain the doctrine and review existing studies, highlighting opportunities to expand the literature.

Google Inc. v. Equustek Solutions Inc.

In Google v. Equustek, the Supreme Court of Canada ordered Google to delist all websites used by Datalink, a company that stole trade secrets from Equustek, a Canada-based information technology company. Google had agreed to do so in part, but with respect to searches that originated from google.ca only, the default browser for those in Canada. Equustek however, argued the takedowns needed to be global in order to be effective. It thus sought an injunction ordering Google to delist the allegedly infringing websites from all of Google's search engines—whether accessed from google.ca, google.com, or any other entry point. Google objected. The Canadian Supreme Court, along with the two lower Canadian courts that considered the issue, sided with Equustek (para. 54). The ruling sets up a potential showdown between Canadian and U.S. law and raises critically important questions about the appropriate geographic and substantive scope of takedown orders, the future of free speech online, and the role of intermediaries such as Google in preventing economic and other harms.