Call for Application

High-Performance Computing Center Stuttgart, Germany (HLRS)

3th – 5th of August 2022 

Organized by the research group “Trust in Information”

Trust in science is of central importance for democratic societies, which aim at participatory decision-making processes. To enable well-informed public discussions and to promote acceptance for political decisions on this basis, scientific results must be widely trusted. However, various forms of mistrust and skepticism about science, as well as confusion about scientific results, seem to be on the rise. They are thus increasingly becoming a social problem.

A key challenge in fostering trust in science is to adequately communicate intra-scientific practices. Science is a complex endeavour in many respects, which is why the interpretation of scientific results is challenging even for scientists themselves. Consequently, uncertainty and profound debates characterise the inner-scientific discussion. However, if the interpretation of scientific results is challenging and already controversial within the scientific community, how are they to be communicated to the outside world without generating mistrust?

The summer school will address this complex of problems from various angles. We aim to better understand the inner-scientific complexity and discuss the question of how it should be dealt with in the interplay between science, politics, media and the broader public. Central questions to be addressed are for example: What makes the interpretation of experimental results so complex that they are usually contested? How should science communication deal with the uncertainty of scientific results and the fact that they are disputed among scientists without causing mistrust? How can science communication reach audiences that are notoriously neglected?

In more detail, we plan to hold the following sessions, of which you can find a more detailed description below:

1. Scientific Practices (Day one)

1.1 Trust in Experiments

1.2 Understanding the Failure of Replication in Science

2. Science Communication (Day two)

2.1 Communication of Uncertainty in the Interplay between Science and the Public and its Connection to Trust        in Science

2.2 Trust-building Science Communication: How to Include Neglected Audiences?

3. Science & Politics (Day three)

3.1 Scientific Disputes

3.2 From the Scientification of Politics to the Politicization of Science

In order to apply for the event, please write to phil@hlrs.de. Please include a short motivation letter (no longer than 1 page) and a short CV. If you are working on a project or a use case that is related to the conference‘s topic, do not hesitate to indicate this as well. The deadline for registration is the 1st of May. It is planned that the conference will take place at the HLRS in Stuttgart. Note that there will be no participation fee.

If you have any questions, please contact phil@hlrs.de.

About us:

HLRS is a research institute and a supercomputer center with one of the fastest computer systems worldwide. HLRS conducts its own research in the field of high-performance computing. Emphasis is placed on the topics of scalability, performance optimization, big data, green IT and the application areas of health, environment, energy and mobility.

HLRS houses a department for the philosophy of computational sciences headed by Dr. Andreas Kaminski. The project “Trust in Information“ is kindly funded by the Ministry of Science, Research and Arts Baden-Württemberg.

Session descriptions

1.1. Trust in Experiment(s)

In this session we will delve into the gritty details of experimental practice. Textbooks often gloss over the messiness of actual data collection and present us with clean graphs from which natural laws seem to follow effortlessly. Even history of science oftentimes traces only successful experiments while dead-end inquiries are only used to illustrate spectacular failures or fraud. Yet the real practice of science arguably contains more stalemates than successful studies, in fact, it even contains more muddy data than is obvious to non-specialists. Nonetheless science progresses from these data. Looking at actual case studies we will try to understand how data is collected and more importantly curated in experimental practices and how this process changes trust.

Questions we want to address in this session are:

1. Is there a specific culture of trust with respect to experiments?
2. Are experiments to be trusted over theories?
3. Can Experiments increase trust in theories and vice versa?
4. What makes one experiment more trustworthy than another?

1.2. Understanding the Failure of Replication in Science 

We tend to take replicability as a reliability indicator of good science. The current so called replication crisis, especially in the fields of psychology and medical research seems to question trust within science (and also its relation to society and the public). Current discussions of low replication rates consider one of the following possible causes: Most focus is put on questionable research practices (QRP) but there are also discussions that the traditional significance level of 5% is too large, that low statistical power overestimates the effect size and so the reliability of findings and eventually that low base rate of true effects strongly influences the replication rate.

Independently from statistical reasons for a failure of replication, in particular with computer based science (e.g. climate science), the influence of IT infrastructures opens a different field of reproducibility.  Here we are confronted with complex and possibly opaque computer programs, huge amount of data and numerical simulations on high performance computer systems. These characteristics of computer based research infrastructures introduce different levels of reproducibility each of which do pose different challenges to yield reliability of results.

1. What do we know about the relative contribution of different factors of QRP‘s?
2. What are the metascientist’s suggestions to remedy the diagnosed flaws in scientific practice?
3. How to reconcile the suggested changes with possible trade-offs in scientific research (Due to limitations on researcher’s control over the research process, they cannot optimize over all aspects simultaneously but must come to terms with trade-offs among desired outcomes)?
4. What are the different levels of reproducibility in computer based research infrastructures? How are they related to both the the mathematical and technical nature of (high performance) computers?

2.1. Communication of Uncertainty in the Interplay between Science and the Public and its Connection to Trust in Science

Climate change or the pandemic have accentuated the presence of scientific results in the public sphere. Uncertainty and catiously taken results are core features of the scientific enterprise. For inner scientific purposes, it is often beneficial to communicate uncertainty since it may enable stimulating collaborations with other research groups or increase the probability of further research funding.

However, these uncertainties are not communicated in the same extent to the public and other key stakeholders. The lay public normally gets not informed extensively about, for example, the generalization power or the limitations of the experimental design like it is the case in the inner scientific discourse.

One reason for this might be that communicating scientific uncertainty may damage the value of scientific knowledge itself and thereby undermine the authority of scientists and other producers of knowledge. Recent societal developments like for example the fabrication of (an allegedly) scientific dissensus and the loss of the internet’s capacity to be a tool for the democratization of (scientific) knowledge does not seem to encourage more openess about uncertainty. So in the so called „Post-Truth Era“ a reasonable assumption may be that communicating uncertainty will reduce public trust.

In this session we want to ask the following questions:

1. How do scientists and science communicators actually cope with the assumption that people may attribute uncertainty to poor science and may even decrease public trust in science?
2. What do we know from empirical research whether communicating epistemic uncertainty across different topics (such as climate change, Covid, migration,…) influences public (mis-)trust?
3. On a methodological level: how do we meassure (epistemic) trust and trustworthiness in an empirical setting?

How can we better understand the relationship between trust in the scientist as an epistemic authority and the persuasive power of the message?  Uncertainty communication might not necessarily raise both at the same time. What can practitioners tell us from their work?

2.2. Trust-building Science Communication: How to Include Neglected Audiences?

A central challenge of science communication is to effectively address notoriously neglected audiences. The typical audience that science communication tries to reach consists of scientifically educated people who are already interested in the subject matter. The flipside is that large segments of the population remain excluded. However, making scientific knowledge and debates widely accessible to many parts of the population, as well as building trust in scientific expertise and experts’ recommendations for action, is essential to enable inclusive and well-informed participation. It is therefore of the essence for democratic societies.

The reasons for the exclusion of certain audiences are manifold. What plays a key role, though, are barriers that are based in the identity and guiding value system of certain population groups. This also manifests itself in an emotional and habitual distance to traditional science communication. This is where science communication can make a difference by trying out new modes and styles of addressing. The workshop therefore aims to better understand the reasons and characteristics of exclusion and tries to explore and discuss ways to respond to it from the perspective of science communication and science journalism. The following questions will be addressed:

1. How, from a science communication/science journalism perspective, can typically neglected audiences be reached?
2. Are there possibilities to address diverse value systems and to justify scientific recommendations with reference to them?
3. What kind of writing styles can make science more accesible? What are the advantages or problems of certain writing styles? Should a more dialogical form of communication be pursued overall, and if so, how?
4. Is the appeal to certain emotions helpful or rather counterproductive to reach to certain groups? Which emotional is suited to appeal to a broader audience without at the same time scaring off others?

3.1. Scientific Disputes

“Trust in Science!“ was the slogan issued during the COVID-19 pandemic in the face of severe and sometimes malicious attacks on scientists. Understandable as it may have been in that situation, it is itself a problematic reduction. Our image of science and its progress is often based on the assumption that dissent primarily concerns the early and surmountable stages of research. The ideal and eventually achieved state is scientific consensus. Issues of dispute are thereby excluded from the social dimension and dealt with in the factual dimension alone.

 Regardless of how accurate this picture is, it makes it difficult for the public to understand and evaluate disputes in science. If scientific disputes do not follow the pattern of being quiet and temporary, then the impression can arise that the sciences do not deliver what they promise. Instead of differentiating the image of science and thus refining our judgment, a withdrawal of trust in science may be the consequence.

 Thus, if one wants to avoid this, the question arises: how should disputes in science be handled? If one is to take the degree of maturity of a science as a yardstick, then not only does the problem arise that the extent of consensus is often taken as an indicator of this degree of maturity. More problematically, dissent is understood primarily as a defect — not as a possible epistemic enrichment of perspective. Furthermore, what about situations where a quick scientific judgment is needed because time is of the essence? Finally, how should outliers be dealt with? Impressions of them quickly swing between scientific revolutionaries and irrational trolls, as it were.

 Furthermore, the thematization of the scientific dispute is often framed by the media. Sometimes, however, this seems not only to duplicate the dispute (it is now carried out in science and in the media reporting on science), but it also gains its own, media-typical dynamics, as can be experienced particularly clearly in the Corona crisis (In the left corner scientist x, in the right corner scientist y. Let’s get ready to rumble!). 

The following questions will be discussed in this session:

1. How should the public, which cannot have the same scientific education in all special fields, come to a judgment how the respective dispute is to be evaluated?
2. How can plausible, legitimate, scientific outsider positions be distinguished from those that do not (should not) have credibility? 
3. How should scientific controversies be presented?
4. How should the relationship between media and science be understood? And how should it be shaped?

In the session “Disputes in the Science” we want to address the triangle of science, public and media. We will analyze the forms of disputes, their temporal course as well as forms of communication and the necessary power of judgment.

3.2. From the Scientification of Politics to the Politicization of Science

In the 19th and 20th centuries, the link between science and politics has been immensely strengthened and multiplied. The social sciences, with their studies of social problem situations as well as sociostructural processes, contribute quantitatively and qualitatively to observing society and orienting political governance. The engineering and natural sciences are changing our lives and forms of practice in a variety of ways; and policymakers are seeking to influence this development through appropriate research funding programs. Model-based research is intended to inform policy decisions in the form of forecasts, for example. At the beginning of the 21st century, even philosophy and theology are at the interface of politics and science in the form of the numerous ethics committees.In addition, there are institutions such as technology assessment, whose founding mission is to provide “policy advice” using scientific methods.

These close couplings between politics and science have also been described as a scientification of politics as well as a politicization of science. Because the two domains are commonly understood as fundamentally distinct from each other (science establishes what is; politics shapes how it should be), the close coupling also raises questions about trust in science.

Questions we want to discuss include:

1. What are the different types of interfaces between science and politics?
2. Does the so-called scientification of politics, for example, increase trust in political decisions?
3. Or does the politicization of science lower trust in the latter?
4. Is it at all appropriate and accurate to speak of the scientification of politics as well as the politicization of science?