Epistemic Networks and Bad Believing: A Workshop in Social Epistemology (16-17 Nov)

Description

We live in the age of strange conspiracy theories, fake news and “alternative truths”: from Pizza-gate and 5G conspiracies, to denying the results of properly organized elections and climate skepticism, many people seem to accept as true claims that are in no way substantiated by evidence. Following Neil Levy, we could call these bad beliefs.

Why do people hold bad beliefs? While it is customary, in traditional analytic epistemology, to study the individual features of people (their reliability, their epistemic virtues or vices etc.) in order to understand how they arrive at knowledge or end up with bad beliefs, recent work in social epistemology has pointed to the importance of epistemic networks in elucidating this matter. People’s beliefs, especially when it comes to social and political matters that pertain to their identity, are often shaped by the social networks the individuals are embedded in.

In this workshop, we aim to discuss the way bad believing (i.e., believing in scientifically unsubstantiated claims that concern vaccines, climate change, migration etc.) is shaped by the epistemic networks of the believers. Some of the questions of interest are:

• Is bad believing a consequence (also) of the structure of the epistemic network individuals are embedded in?
• What are the structural properties of the networks (diversity of sources, independence of informants etc.) that might be determinant for why people end up with bad beliefs?
• How to characterize epistemic vulnerability in epistemic networks?
• Are there differences in the structure of the epistemic networks in digital spaces as opposed to non-digital spaces? And how can the former be studied?
• How can epistemic networks be modified so that to avoid bad believing?
• What are some of the individual virtues that would help protect against harmful configurations of epistemic networks?
• How to conceptualize the role of trust in epistemic networks?
• How can expertise be incorporated in epistemic networks so that to ameliorate epistemic harms?
• What is the role of emotions and other affective states in shaping people’s attitudes towards different sources of information from their epistemic networks?
• What other non-epistemic factors are relevant for understanding network influence?
• What lessons (for ameliorating lay-people bad believing) can be extracted from the study of scientific networks?
• How to conceptualize epistemic network injustice (i.e., epistemic injustice that is due to harmful epistemic network configurations) and what tools are there to ameliorate it?

While recent research on epistemic networks comes especially from philosophy of science and from the study of communities of scientists, our hope is that the study of epistemic networks can be helpful for understanding the epistemic behaviour of non-scientists as well. The calls for understanding the mechanisms of bad believing in the political domain (and the ameliorative tools for overcoming dangerous political beliefs) are more and more urgent. As many have pointed out, the health not only of our democracies, but of the climate itself might be at stake.

Location

This is a hybrid event (all talks are also on Zoom).

In person, at Open Science Lab, room OSL 3, Zellescher Weg 21-25,  01069 Dresden. Posters and signs will guide you to the room (OSL 3 is on the first floor, at the end of the long corridor on the right).

Digitally, via Zoom, upon registration. 

Program

Thursday (16^th of November)

08:30 – 09:00 >     Welcoming Coffee and Registration

09:00 – 10:30 >     Mark Alfano (Macquarie University) - Personality, social network, and natural language approaches to intellectual humility and its correlative vices

10:30 – 11:00 >     Coffee Break

11:00 – 12:30 >     Mandi Astola (Delft University of Technology) - Mandevillian Vices in Epistemic Networks

12:30 – 15:30 >     Lunch Break and Exhibition at TU Dresden

15:30 – 17:00 >     Dunja Šešelja and Matteo Michelini (Ruhr University Bochum) - The Misinterpretation Effect: How Science Benefits from Evidence Misinterpretation

17:00 – 17:30 >     Coffee Break

17:30 – 19:00 >     Cailin O’Connor (University of California, Irvine) - Virtues, Vices, and Networks: How networks transform the epistemic impacts of individual behaviors

20:00 – 22:00 >     Dinner

-----------------------------------------------

Friday (17^th of November)

08:45 – 09:00 >     Morning coffee

09:00 – 10:30 >     Sergiu Spatan (Dresden University of Technology) – The Role of Strong Ties in Holding (and Avoiding) Bad Beliefs

10:30 – 11:00 >     Coffee Break

11:00 – 12:30 >     Patricia Rich (University of Bayreuth) and Emmanuel Genot (Lund University) - Mutually Expected Rationality in Online Sharing 

12:30 – 14:00 >     Lunch Break

14:00 – 15:30 >     Sven Engeßer (Dresden University of Technology) - Climate Change and Social Identity

15:30 – 16:00 >     Coffee Break

16:00 – 17:30 >     Thomas Grundmann (University of Cologne) - Epistemic Nudging as Antidote to Bad Beliefs

Abstracts

Mark Alfano - Personality, social network, and natural language approaches to intellectual humility and its correlative vices

I review several years' worth of studies of intellectual humility and its correlative vices by my interdisciplinary research group, using approaches from personality and social psychology, social network analysis, and natural language processing. In the context of personality, we find that intellectual humility and its correlative vices consistently cleave into five dimensions: two virtues (individual openmindedness and ingroup criticism) and three vices (sense of superiority, intellectual arrogance, and emotional defensiveness). We find that ingroup criticism and arrogance are especially predictive of a range of relevant attitudes and behaviors, such as acceptance/rejection of warranted/unwarranted conspiracy theories, medical (mis)information, and social feedback on objective tests of reasoning and intelligence. In the context of social networks, we find that different networks are better or worse structured to enable their members to benefit from the wisdom of crowds, which is possible when one has multiple, independent, diverse sources that are more reliable than chance. And in the context of natural language processing, we find that people who reflect on their experiences using self-distancing techniques both express more intellectual humility and use fewer words associated with anger and the present tense. Together, these studies shed light from multiple angles on the ways intellectual humility and its correlative vices are manifested in everyday life.

Mandi Astola - Mandevillian Vices in Epistemic Networks

Mandeville’s famous work “The Grumbling Hive” puts forward the argument that individual vices can sometimes be valuable for society; egoistic striving ensures economic growth and welfare. Individual vice, whilst detrimental to individual flourishing, can play a part in constituting collective flourishing. More recently, this idea has been applied to the epistemic activity of groups. Some have argued that epistemic vice can, in some cases, contribute to collective epistemic flourishing. This has sometimes been called Mandevillian virtue or Mandevillian intelligence. However, Mandevillian virtue also has a negative counterpart: Mandevillian vice. Mandevillian vice is when virtuous character traits of individuals contribute to collectively suboptimal epistemic behaviour. An example of this would be when virtuous belief adoption under conditions of uncertainty causes a lack of viewpoint diversity. In this talk, I outline the possibility of Mandevillian vice in epistemic networks, and its implications for virtue epistemology.

Dunja Šešelja and Matteo Michelini - The Misinterpretation Effect: How Science Benefits from Evidence Misinterpretation  (joint work with Wybo Houkes, Javier Osorio and Christian Strasser)

In philosophy of science and social epistemology, it is widely agreed upon that not responding correctly to evidence is epistemically deficient. Failing to respect one’s evidence includes cases such as ignoring data, misinterpreting it, or jumping to unwarranted conclusions. While this point may be indisputable at an individual level, the situation is less clear at a collective level. Our paper questions whether a group of agents, including those who misinterpret the available evidence, should always be considered epistemically inferior to a group without misinterpreters. We argue that this is not necessarily the case by presenting a model-based argument against the claim that individual failures to respect evidence necessarily lead to harm at the collective level. More specifically, we use an agent-based model to show that communities in which some scientists misinterpret some of the evidence may outperform communities in which no single scientist misinterprets the available evidence. We call this effect 'Misinterpretation Boosts Accuracy' of inquiry. This effect is a specific manifestation of the ‘Independence Thesis’ (Mayo-Wilson et al., 2011), according to which norms guiding individual rationality may not be optimal for the epistemic performance of a group, and vice versa.

Cailin O’Connor - Virtues, Vices, and Networks: How networks transform the epistemic impacts of individual behaviors

In an age riven by misinformation, much new research has focused on the role of intellectual virtues and vices in the spread and persistence of false beliefs.  This talk uses a series of network models to explore how individual virtues and vices might function in a group setting.  The picture that emerges is one where certain epistemic vices, especially those related to intransigence, can benefit groups.  This lends credence to previous accounts pointing to disconnects between epistemic goods at the individual and group levels.

Sergiu Spatan – The Role of Strong Ties in Holding (and Avoiding) Bad Beliefs

Recent research in social epistemology has focused on the structural features of people’s epistemic networks to explain why they might end up with false and unsubstantiated beliefs. In this framework, it is often argued (see especially recent work by Emily Sullivan and colleagues) that the number of one’s sources, their diversity and their independence are key structural properties determining one’s epistemic position in a network. In this talk, I take another route and look at the role of strong ties (e.g., friends or relatives, as opposed to weak ties, e.g., mere acquaintances) in shaping people’s beliefs. Inspired by recent sociological research on the strength of strong ties (see Damon Centola and colleagues), I argue that with some topics - those relevant for one's social identity - it matters very little how diverse or independent one’s sources are, but much more how reliable one’s strong ties are (with regard to those topics). This framework, I claim, suggests interesting strategies for dealing with the current influx of bad believing.

Patricia Rich and Emmanuel Genot - Mutually Expected Rationality in Online Sharing 

Models of content-sharing behavior in online social media platforms are studied at two levels: At the content level, content is represented as diffusing, as with contagious diseases. At the user level, individuals are represented as making choices as in single-agent decision theory. Social media platforms are interactive spaces, however, in which agents care about what their fellow agents post, share, like, and dislike. This means that the agents must reason about how others will respond when choosing their own actions. Existing models leave this strategic reasoning out, but we show that it has important consequences. We model agents' social media decisions using a cognitive hierarchy framework. Analytically, we show that there are limit cases in which the platform can be swamped with content that no agents personally like (think obvious fake-news). We then use agent-based simulations to explore more realistic cases which can give rise to similar outcomes.

Sven Engeßer - Climate Change and Social Identity

There is still the widespread notion that climate change contrarianism results from a lack of awareness and knowledge. However, empirical studies could not find any support for this hypothesis. Individuals with higher degrees of formal education and science literacy were not the most concerned about climate change. Among them, cultural polarization was the greatest. There is indication that cultural influences and identify protection are more meaningful than rational thinking, culminating in the “white‐male effect” (Kahan et al., 2007; McCright & Dunlap, 2011). Communicators should take these findings into account. 

Thomas Grundmann - Epistemic Nudging as Antidote to Bad Beliefs

Bad beliefs are radically detached from truth. They may result from massive disinformation, structural insulation from relevant evidence, or the agent’s enduring unresponsiveness to evidence due to ideology, polarization, or bias. In my talk, I will focus on the case of evidence-unresponsiveness. As I will argue, nudging can be used to restore not only true beliefs but also knowledge in people who are unresponsive to evidence. My argument relies on some controversial assumptions about nudging (as by-passing evidence) as well as knowledge (as not requiring evidence and relying on externally individuated cognitive processes). In my talk, I will defend these assumptions and explore more broadly the prospects and problems of epistemic nudging.

Organization

Organization team:

• Moritz Schulz;
• Sergiu Spatan;
• Klara Gaßner.

Part of the project „Gefangen im Netz? Chancen und Perspektiven von epistemischer Netzwerkanalyse“, financed by Sächsische Aufbaubank, and based at Technische Universität Dresden.