Third Narrative Science Seminar: The functions of failure narratives in engineering, and narrating causality within experimental psychology
Author: Dominic Berry
So far our seminar series has guided us through cases of narrative science in synthetic biology, climate policy, philosophy of modelling, and economics. The speakers in our third seminar, Caitlin Donahue Wylie and Sigrid Leyssen introduced two additional areas, respectively engineering and psychology. As with all of the blog posts following these seminars, they are designed to capture some central elements of the presentations and audience interaction, and are not by any means a complete record of the event.
Caitlin Donahue Wylie (University of Virginia)
Narrating Disaster: A Method of Socialization in Engineering Laboratories
At the outset Wylie explained how this paper concerns a new project, one that builds on her previous work investigating the knowledge-making of laboratory technicians. Where once she had investigated lab technicians, recovering and unpacking their contributions to cultures of science and to knowledge, she is now instead interested in STEM students, though very much with the same kinds of question in mind. Narrative enters her picture thanks to extended periods of laboratory ethnography, in which she finds engineering educators, tutors, and lab assistants, making habitual use of narrative for diverse ends. Some of these are pedagogical, but others are equally about engineering identity: what does it mean to be a good engineer? The instances in which narrative manifests are often motivated by epistemic uncertainty, how to understand uncertainty, handle it, be prepared for it, and so on. In recognising the uses of these kinds of narrative she is reminded of Sharon Traweek’s Beamtimes and Lifetimes which includes, amongst many other things, an in depth analysis of ‘male tales’: the kinds of story told and shared ritualistically amongst male physicists that are intended to solidify a common understanding of what scientific excellence is, and who are the kinds of people that can be expected to attain it. Wylie certainly found partial evidence of male tales in her labs, but more importantly, most of the narratives being told and retold in engineering education spilled over these categories.
Through her examples Wylie then established and argued for the importance of ‘vicarious learning’, a form of instruction that involves a teller and a listener, or listeners. This relationship asks the teller and the listener(s) to enter into a moment of learning that is not simply an exchange of knowledge, or an instruction, but makes use of the tellers’ having once been in the listeners’ shoes. These moments are intended either to show the listener ‘what has always happened before’ or, ‘the mistake I always make is’, or ‘you have to watch out for’, and so on. In particular she finds failure or ‘disaster’ narratives particularly common in instances of vicarious learning, and here it matters that narratives are able to carry or build sympathies and normative assumptions along with shared understandings. Vicarious learning is fundamentally constituted by narrative because it recognises all of the participant’s auto-histories, how they have come to reach this moment, without stating those terms explicitly, precisely as we on the Narrative Science project have come to expect in the making of ‘tellable’ narratives.
In the question period the audience asked about the aspect of this kind of narrative that works to build a team rather than individuals, and also the extent to which narratives of disaster and failure might actually be quite different. It might be, for instance, that a disaster narrative builds a foundation for later greatness, overcoming something so terrible. Whereas failure narratives might just be a let down. Other participants asked whether the sharing of failure narratives actually had some downsides. Wylie could not think of any immediate negative effects, and could instead point to how in one of the labs where she has been conducting an ethnography, on hearing about her research, the students spontaneously decided to make a ‘disaster story shelf’, which showcases the results of their mistakes (broken beakers etc.). Another person wondered whether or not failure narratives were also inuring students to failure in large technological disasters, as something that engineers have particular authority over and are expected to master, rather than something that might require humility. This would be a case where engineering epistemology meets identity.
Sigrid Leyssen (Bauhaus Universität Weimar)
On the Experimental Phenomenology of Causality
Leyssen’s talk likewise took some of her previous work as a jumping off point in order to launch her brand new project on the history of the scientific experiments of the filmology movement, and experimental phenomenology, narrative providing a persistent feature of interest throughout. She began by looking at cases in the history of science where the fact of narrativity itself was the research object for scientists, in particular for experimental psychologists. Several psychological experiments could be read as studying narrativity in some of its most basic forms, for example, as merely causal relations between events. Nevertheless, the diverse terms on which someone might choose to understand something so apparently simple soon demonstrated the complexities of narrative knowing. Her primary case study was the work of Albert Michotte, which, starting in 1939, sought to find evidence of causal understanding directly in the perception of events, rather than as the result of a higher order cognitive process: narrative, or causation, as something that one just knows from witnessing it. For these purposes Michotte created a range of different experimental apparatus intended to prove his point, which he would travel with and demonstrate at international conferences. Some of the most charismatic include spinning wheels with different patterns, the motion of which seemed to imply a causal relation between the elements of the patterning. For an example of this kind of effect, see the following video.
Other contemporary examples of psychological experiments that could be interpreted as working on the narrativity question as a psychological problem are the investigations of Fritz Heider and Marianne Simmel into how and why people read anthropomorphic narratives into abstract moving objects and shapes. For an example of the kind of movie they created for these purposes, see:
She also mentioned the earlier work of Frederic Bartlett – a good friend of Michotte – on memory. Bartlett worked with stories as his experimental material, studying where stories changed when people were asked to recount them.
In the second part of her talk Leyssen focussed on how Michotte’s work was taken to matter for scientists, philosophers, and filmmakers more broadly. Only able to bring his work to an international audience after the end of the Second World War, by this point Michotte had also become interested and active in the ‘filmology’ movement. Research questions in this emerging field included ‘how can people see events in fragments? What are the minimum number of gaps, or the maximum number, that enable or interrupt the perception of relations of temporality, spatiality, causality?’. These results mattered for psychology but also for those experimenting with film making and art, so we here have a history of narrative scientised and literised at once.
In the question period we focused on the kinds of controversy that surrounded Michotte’s arguments, whether, for instance, the argument that narrative works by compulsion was disagreed with. Leyssen explained that many of Michotte’s views on the perception of causation were heavily criticised, and that debate as to whether causation is based in recognition or perception continues today. Meanwhile Michotte is having a renaissance, many researchers in the past ten years taking up his assumptions and research programme. Another participant wanted to hear more about the context of the growth of phenomenology at this time, and if there are further examples of people investigating perception on these terms. While referring to the vibrant surge in philosophical phenomenology at the time, also at Michotte’s home institution, the philosophy institute in Louvain (where the Husserl archives had just moved), Leyssen could also point to the earlier work of Max Wertheimer, and the broader legacy of Gestaltist researchers in the parallel tradition of experimental phenomenology in the scientific psychology. This had its own impact throughout the sciences and arts. Attending film festivals, and attending to the structure of narrative there, was something that Michotte and others like him gave increasing attention to over the course of the century.
The final seminar of Term 1 took place on the 4th of December, and featured Lukas Engelmann on ‘Epidemiology as Narrative Science: Outbreak reports of the third plague pandemic from 1894 to 1952’ and Sabine Baier ‘How Many Molecules Does It Take To Tell A Story?- Managing Epistemic Distances In Medicinal Chemistry’. A blog post will follow soon!
We also hosted our latest workshop on the 10th of December on ‘Expert Narratives: Systems, policies, and practices’ organised by Dr. Mat Paskins and Prof. Mary S. Morgan. We are sorry to all those of you who applied to attend but missed out this time, numbers were necessarily limited. We have further workshops and conferences planned for 2019 so as ever, join the mailing list to keep up to date. A blog post following the Expert workshop will also be published here soon. Our considerable thanks to all the speakers and contributors to discussion!
Second Narrative Science Seminar: Tidal Narrative Ordering for Whewell and San Francisco, and Independent Discovery of Utility Narratives in Economics
Author: Dominic Berry.
These two talks furthered the Narrative Science seminar series, which will continue throughout 2019. For more details please see the Events page of the website. The first paper was presented by Julia Sánchez-Dorado and Claudia Cristalli, both of UCL, working in the mode of integrated HPS. The second was given by Prof. Mary S. Morgan, leader of the Narrative Science project, based on a presentation given a week earlier in Seattle at the History of Science Society annual conference.
Julia Sánchez-Dorado (UCL) and Claudia Cristalli (UCL)
Colligation in model analysis: from Whewell’s tides to the San Francisco Bay Model
One aim of the talk was to bring attention to the first philosophical articulation of ‘colligation’ by William Whewell, an idea that has been part of the bedrock of the Narrative Science project from the getgo, and which can be found in a number of articles in the 2017 ‘Narrative in Science’ special issue. They wished to preserve particular aspects of Whewell’s original conceptualisation which might otherwise be lost if we broaden the notion to simply mean ‘gathering things together’. Colligation as described in Whewell’s The Philosophy of the Inductive Sciences (1847) arises as an ‘act of thought’ of bringing together empirical evidence and placing a novel conception onto them and their relations. Key features were then drawn out through Whewell’s work on tidology, the study of tides, for which Whewell thought colligation was essential, as without it one might never reach a systematic and law-like account of tidal phenomena. His state-of-the-art was to study times of low and high tides for different coasts around the world, as then known, to draw ‘co-tidal’ lines between points where low and high coincide, in the process finding global tidal relations. This was colligation in action, and for Whewell, was a way to begin generalising in a mapping practice what was otherwise a seemingly random phenomena.
Sánchez-Dorado and Cristalli then compared this kind of data collection and modelling with a C20th example, engineers and scientists trying to model tidal activity in the San Francisco Bay. This model was meant to capture the tidal activity of the area accurately enough that it could be used to predict the future effect of, for instance, building a damn ahead of the bay. But how to select the relevant features to model, and how to relate them? Here they find that colligation was again central, not only in deciding which features mattered for the model, but also for assessing how well the model mapped onto the behaviour of tides found in the Bay.
In the question period we focussed on different notions of colligation, for on the one hand the making of a series of observations and their compiling in a model, and on the other the use of the model to answer particular questions. The extent to which narrative mattered in these two senses could be clearer, and more work could be done to draw out the importance of narratives for using models as opposed to making them in the first place. Others wanted to know more about the process of using colligation in producing things like co-tidal lines, where idealisms and visuality come into what Whewell is producing. In response we heard that there were two primary competing theories of tides in the C19th, the one that argued tides worked synchronically - i.e. independent of one another - the other that the globe should be thought of as essentially a large box full of water which sloshes about creating patterns of regularity. Colligation mattered in particular for bringing together all the high tides to create a visual picture of patterning, in the process exaggerating and idealising the lines he ‘saw’, but this was only one part of Whewell’s search for regularities.
Mary S. Morgan (LSE)
Simultaneous Discovery or Competing Concepts? Economists's Notions of Utility in the Late 19th Century
This paper focussed on narratives in concept formation, Morgan deriving evidence of multiple kinds of small or large narrative being involved in the making of concepts of utility used by C19th economists. Kicking off slightly earlier, Mary emphasised how Adam Smith’s argument that wealth creation is the result of divisions of labour was effectively a large narrative about the interrelations of states. Smith’s labour theory of value became one of the most important and influential narratives available to C18th economists and well into the C19th. Labour was seen as a key constituent of what makes use value and exchange value, which are consistently explained through short simple and idealised real-life examples, such as use of labour in the hunting of deer and bears and the exchange value of their hides relative to the labour it takes to kill them. Smith’s ideas would be overtaken in the late C19th by the notion of ‘marginal utility’ established independently by at least four different economists working at this time; Carl Menger, John Bates Clark, William Stanley Jevons and Léon Walras. Their publications are classical foundational texts for modern economics. With marginal utility what matters more is the extent to which any given person ranks the value of a good for themselves, which may or may not have anything to do with the amount of labour that the good embodies.
Morgan points out how each of these authors attempted to establish marginal utility as a concept, often relying on narrative for these purposes. For example Jevons’ utility curve measures the increase or deterioration of a good’s desirability as more is consumed, which as with the earlier generation of economists, he illustrates through short literary accounts. One example is the ‘paradoxical parable’ of the value of water in contrast with diamonds, which he explains as not actually problematic or paradoxical because there is loads of water available but diamonds are scarce and produce a different experience for the acquiring consumer. In Jevons alone there are other narratives of human behaviour, which he mathematizes, and further thought experiments regarding, for instance, how to decide the apportioning of food on a ship lost at sea. A great many more narrative examples can be found in all of these authors, all of which Morgan argues contributed to the making and refining of concepts of utility.
In the question period the audience wanted to know more about the range and functions of the short narratives used by economists to illustrate differential ranking of need and value. Did they, for instance, ever reflect on the normative or moral components of these short real-life examples? Morgan has yet to find any evidence of this kind of reflexivity! Morgan also took the opportunity to clarify a clear difference between the earlier examples associated with Smith, which are clearly designed to draw the reader in and have a lot of rhetorical power, whereas the later examples from Jevons et al. which tend to get more intricate and are focussed on how to think through a given problem. Others asked about the extent to which all of these little narratives are doing different kinds of work, and whether any two economists alighting on a particular narrative necessarily means they are talking about the case for the same reason. In response Morgan points out how widespread were some of these examples, which could be seen as forming a crucial part of the discursive element of the development of economic theory. Robinson Crusoe for instance is picked up and applied by most economists working at this time. Nevertheless she agrees that the uses to which such narratives were put could vary widely.
We have already hosted the third Narrative Science seminar, featuring Caitlin Donahue Wylie and Sigrid Leyssen, about which a blog post will soon follow.
The next Narrative Science seminar, which will be the last one for this term, will take place on the 4th of December and features Lukas Engelmann on ‘Epidemiology as Narrative Science: Outbreak reports of the third plague pandemic from 1894 to 1952’ and Sabine Baier on ‘How Many Molecules Does It Take To Tell A Story?- Managing Epistemic Distances In Medicinal Chemistry’. As ever, please see the Events page for their abstract and information about timing and location.
We have also recently announced a call for applicants to attend our next Narrative Science workshop, Expert Narratives: Systems, policies and practices. Spaces are limited but we will attempt to accommodate as many of you as we can. You should also know that we have made available 4 PhD student travel bursaries, details for applicants are all found on that same workshop notice.
We have just announced details regarding our next workshop. Everything you need to know regarding how to apply to attend, and also details for how to apply for the available PhD student travel bursaries, is included in the Events page under the section 'Upcoming workshops': https://www.narrative-science.org/events-narrative-science-project.html
Places are limited but we will try and accommodate as many of you as we can.
Our next post will offer a summary of the presentations given at our last Narrative Science seminar, so watch this space!
First of the Narrative Science Seminars: Synthetic biological beings and the narrative futures of climate science
Author: Dominic Berry
Earlier this month we launched the Narrative Science seminar series. We were very fortunate to host two outstanding researchers, Dr. Sally Atkinson and Dr Elisa Vecchione, working respectively at the intersections of narrative and social anthropology, and narrative and international governance. This post gives a very brief overview of key aspects of their talks, illustrating moments where they directly informed or provided challenges for our project.
Sally Atkinson (University of Exeter)
Fragile cultures and unruly matters: narrating microbial lives in synthetic biology knowledge practices
Atkinson set out to better understand the knowledge and research practices of researchers developing new techniques in the field of synthetic biology for industrial bioproduction. As part of the DETOX project, drawing on interviews and laboratory ethnography she works closely with researchers analysing and developing synthetic microbial strains. Narrative presented itself to her in these settings in a number of different ways, used by different actors. Some of the most explicit are the highly promissory narratives around environment and sustainability, written into research proposals and policy documents. Her attention is directed towards how such narratives play out in practice in emerging sites of knowledge. She suggested that to meet these promises practices of genetic mapping and multi-omics data analysis are used to construct microbes as a form of technology for bioproduction, made meaningful and suitable for travel and modification. Such data journeys, Atkinson suggested, might be best understood and further explored as having a narrative form.
Through her interviews and observations Atkinson finds synthetic biologists talking about their research materials not only as objects, easily moved and re-designed, but as actants, presented as having their own unruly lives and narratives. She was particularly interested in the tension this indicates between stable models of microbes as promissory technology and the challenges involved in working with microbial matter. Building on the work of Wynne and Strathern she suggested such narrative tension between model and practice demonstrate how researchers continually work to hold together projects in the face of uncertainty. In the discussion period we considered whether or not failure narratives, as a genre or type of narrative, allow for and validate how and when things go wrong in experimental spaces, to what extent is was appropriate or useful to describe such narratives through the lens of care, or whether epistemic problems at hand were about a managing and framing experiences of failure and notions of responsibility.
Elisa Vecchione (Group of Pragmatic and Reflexive Sociology, EHESS, Paris)
The political necessity of a more poetic science: the case of climate-economic narratives
Vecchione began by explaining how her work explores and understands the modelling of climate change, and its impacts on (or failure to impact) international governance law and policy. As a matter of contingency, she was introduced to the work of Hayden White, and immediately recognised in his theorisation of narrative in history a new way to understand climate modelling as creating narratives of the future. She is particularly interested in how climate scientists produce their evidence and then organise it for the purposes of, for instance, planning and assessment within the Intergovernmental Panel on Climate Change (IPCC). Longer term explanations for global climate change are most visible in at the IPCC through their scenario building and reporting exercises. Vecchione emphasised that models at the IPCC are not built in isolation of everything else, but are rather prepared in concert with particular scenarios that lay out the key storylines within which any given set of models need to make sense or be understood. In her talk Vecchione built upon Jerry Ravetz’s work on models as metaphors for these purposes, but closer to the home of Narrative Science, we also recognised a line of argument developed by Mary Morgan which had helped develop the foundations for our current project.
The primary aim of the paper was to demonstrate the variety of different ways in which narrative is central and essential to climate modelling, obviously not with the intention of thereby undermining the value of these efforts. (The idea that the presence of narrative in science demonstrates that science’s immaturity or falsity is a very old prejudice that our project likewise tackles head on). Rather the intention is to make climate scientists more conscious of the importance of narrative argumentation in the process of model building. In particular Vecchione advocated for the importance of retro-dication as a way to assess the quality and viability of different climate change models.
In the discussion period we wanted to know more about how different potential ‘moments’ are included as admissible into a climate change model, and also the ways in which narrative functions in the consciousness of model making, in the making of projections, and also bringing these to the public in a more deliberative democratic model. This resulted in Vecchione thinking more about the particular audiences served by these different narratives, and how very different kinds of narrative, or what we mean by narrative, can be found in these settings.
The next seminar will take place on the 6th of November, featuring Julia Sánchez-Dorado and Claudia Cristalli on ‘Colligation in model analysis: from Whewell’s tides to the San Francisco Bay Model’ and a second speaker we have yet to confirm. We are very sorry to report that our scheduled speaker, Dr. Veronika Lipphardt, has had to cancel this time for personal reasons, but we are working with her to find a suitable alternative time to host her in the future. Instead Prof. Mary Morgan will present her new paper ‘Simultaneous Discovery or Competing Concepts? Economists's Notions of Utility in the Late 19th Century’.
You can also find many of us contributing to the upcoming meetings of the History of Science Society and the Philosophy of Science Association.
HSS (Seattle, 1-4 November)
Friday 2nd November 12:00-13:15 - Round table - 'When stories are science'
Sunday 4th November 9:00-11:00 - Symposium - 'Scientists' narratives'
PSA (Seattle, 1-4 November)
Poster - 'Developing a philosophy of narrative in science'
Friday 2nd November 18:00–20:00 – Poster Session - 'Developing a philosophy of narrative in science' (poster number 52)
Hello and welcome!
We are very excited to begin sharing our research with you through this site. This will be our main way to communicate with you all, keeping you up to date on our work as it develops, the events we are organising and taking part in, while keeping the discussion going through posts and publications.
Elsewhere on this site there is a detailed Introduction to the project aims and ambitions. You can also learn more about who we all are from the Team page, and more on the funding of the project through About. The project was also recently featured in the Viewpoint magazine of the British Society for the History Science.
If you don't want to have to click around (but you really should!!) then a brief introduction below provides a couple of routes into the project. We are hoping to interact with a wide variety of experts on the relations between narrative and science, so one aim of these introductions is to build interest and connections. We would love to hear from you in the comments below, or via email.
What is Narrative Science?
In the course of their activities, scientists often construct and rely upon narratives. Narrative involves ordering materials, an ordering that can be achieved in a variety of ways, be it visually, through diagrams, flowcharts, maps, and the like, or through prose. We can often recognise scientific arguments and practices as adopting or containing narrative structures and elements. What can we learn by subjecting these uses of narratives, their authors, characters and events, to serious scrutiny in order to appreciate the logics and rationales by which scientists’ narratives work? The Narrative Science project explores the philosophical, historical, social, and epistemic functions of narrative in the sciences, and analyses historical cases in which they have been significant.
Another set of questions that we are interested in, are more specifically related to broader questions in the history and philosophy of science. How, for instance, does new knowledge become established, and how do novel categories, ontologies, or epistemologies emerge? What epistemic strategies and tools are available to researchers as they attempt to persuade fellow scientists of the truth of their observations and arguments? What kinds of broader concepts and categories of knowing underpin what counts as successful science? In these respects we are dedicated to finding and demonstrating the ways in which narrative knowing has been central to precisely these kinds of social and epistemic developments throughout the history of the modern sciences.
For more please visit the Introduction.