Annotated Bibliography

This paper is provoked by the rise of a particular kind of game theoretical analysis within the social sciences and history, that of ‘Analytic Narratives’ (ANs). It therefore discusses features of a kind of case ripe for exploration in the Narrative Science project. The author argues that one cannot expect ANs to explain the phenomena that they correspond to, rather their primary usefulness is in the construction of hypotheses that will more or less successfully explain the phenomena. This is particularly interesting from the perspective of Narrative Science, because it reminds us that we may not need to try and find narrative working in isolation of other features of analysis, experiment, explanation, etc. but narrative may instead be a contributing factor in a larger whole. However, the paper also leaves open certain questions that our project would otherwise want to prioritise. The author only addresses the model component of ANs, rather than the model and its narrative. Every AN is composed of a model and its associated narrative, which together contribute to hypotheses of an overall explanation. This author chooses to bracket out the functions of narrative, and instead considers only the ways in which a model may or may not contribute to an explanation of a phenomena. Rather than being a problem for the paper, this provides one obvious line of future research that the Narrative Science project can highlight and encourage. 

LINK TO PAPER

In this paper Ankeny provides an account of the creation and use of case histories in medicine, one that enriches our understanding of their epistemic content and functions. Though often viewed with suspicion due to their describing singular or small sets of patients, the medical case plays a key role in the creation of new diagnoses and the process of matching? new patients with existing diagnoses. These starting points are very sympathetic to the Narrative Science agenda, and indeed it seems clear that Ankeny’s account could be expanded in this direction. As Ankeny writes: 

“In other words, the publication of a case is in some sense a proposal or promise that the particular will be able to be made generic through a process of systematizing the original facts, smoothing out their particularity and refining the original case until it is applicable beyond an individual patient”. 

This whole process, and the ordering and systematizing that goes into turning cases into diagnoses, can be fruitfully interpreted as a narrative practice. 

LINK TO PAPER

What has been the influence of scientific literature on literature more broadly, and when has artistic expression influenced the sciences? Even if we do not often find direct connections, that a historical period may be investigated by considering forms of writing or common languages and metaphors in many domains across the sciences, arts, and humanities, searching for their differences and similarities, is a highly productive and rewarding way to pursue the history and philosophy of science. Gillian Beer’s work launched a multitude of studies at the intersection of history of science and literary studies, helping scholars revise their understanding of each independent of, and in connection with, the other. Not only does Beer’s work provide some of the inspiration for Narrative Science, it also provides a case study that we may wish to return to, but with our own priorities. For instance, rather than looking at the broader impact of Darwin on literature, we might want to assess the underpinning of evolutionary accounts more specifically, to explore in depth when, where, and how Darwin was producing narrative knowledge—partly through his incorporation of metaphors and myths also circulating in literary spaces, but also through his processes of ordering, or in interaction with models, diagrams, and other forms of analysis.

LINK TO BOOK 

At the intersection of the philosophy of history and literary theory, Danto looms large. Within an overall scheme that turns narrative into sequence, with the familiar beginning, middle, and end, particular discussion has surrounded his notion of the narrative sentence, which is no often referred to as the ‘Danto sentence’. These are sentences that possess a particular set of qualities, their presence in a text often serving as a diagnostic that narrative is present. The most important criteria for a Danto sentence, is that the statement contain reference to at least two entities (objects, places, persons, etc.), which have been separated in time, and that the sentence is about the first of the two to have occurred temporally. By ‘about’ Danto means? mean, the first entity is the one for whom the information revealed is the most significant. The earliest articulation of Danto’s argument in this regard can be found in “Narrative Sentences”, History and Theory, Vol. 2 (1962), pp. 146–179. There Danto provides an example of a narrative sentence in action:

“An example, for an historian writing about the events leading up to the outbreak of war in 1914, might be: "And so, at Sarajevo, the first shot of the First World War was fired." Such a statement, although about an event at one time, refers forward, not just to something envisaged by the historical agents concerned, but to the actual subsequent course of events. It is a judgment which could be regarded as typically historical, since it introduces a kind of retrospective intelligibility into the account of what happened. And it clearly does this by connecting facts or events at different times.” 

This emphasis on the connection between facts is one that Narrative Science is dedicated to exploring throughout the sciences, not just within literature and historical writing.

LINK TO BOOK

Forrester’s research provides the most in-depth analysis of the historical emergence and epistemological functions of case-based sciences in the late nineteenth century, with a particular focus on psychoanalysis. This short chapter provides an introduction to themes and research that have been published more comprehensively as Thinking in Cases (2017), Polity Press. Here Forrester is particularly focussed on the ways that case histories in psychoanalysis not only communicate facts about patients, and the relationship between the patient and medical practitioner, but are also cultivate the proper psychoanalytical posture of the reader, and their understanding of psychoanalysis in practice. The particular narrative of the case therefore plays a number of different roles simultaneously. Forrester also demonstrates that throughout the history of psychoanalysis, the emphasis on singular cases was always under threat from those seeing to systematize, or even ‘scientize’, research practices. How and why narrative science flourishes, or comes under threat, provides part of the backdrop to our project’s overall investigation. 

LINK TO BOOK

Aside from establishing an innovative account of what models are and how they work in science, this paper also provides an excellent example of how working through analogies between literary and scientific things can yield important insights into the philosophy of science. As Frigg writes:

“drawing an analogy between scientific modelling and literary fiction is not idle musing; it is the driving force behind an approach to scientific modelling that aims to provide an understanding of a central aspect of scientific practice.”

The specific fictional account that Frigg takes from philosophy of literature and inserts into his analysis of scientific modelling is that of ‘pretence theory’, as developed by Kendall Walton. First setting aside the term ‘model’ as having far too many meanings, the paper moves on to consider the anatomy of a ‘model-system’ and the features of theorising and data gathering that might feed into it. The primary features of a model-system are then likened to those of fictional characters and places in literature, which in both cases do not have to exist in the world in order to facilitate analysis and understanding. Three kinds of fictional statement in particular are attended to—intrafictional, metafictional, and transfictional—all in aid of Frigg’s new fictional theory of scientific representation. As a paper that not only exemplifies what can be gained by looking at philosophies of science and literature together, but also forges links between narrative, imagination, modelling, and representation, it is most valuable to Narrative Science. One introductory section also builds a comprehensive list of earlier efforts to understand scientific models in light of fiction. 

LINK TO BOOK

In this article the philosopher Walter Gallie establishes some of the fundamental ideas that would go on to be published in his highly influential Philosophy and the Historical Understanding (1964). Here we focus on the first section of the article ‘What is a story?’ because it gives us plenty to unpack for the purposes of Narrative Science.

First, Gallie’s chosen concepts for probing what makes a story, and how a story works, invite incorporation into exploration of scientific narratives. ‘Followability’, ‘interest’, and ‘conclusions’, these are all standards, or conventions, or structures, that we also expect scientific work to conform to. Take followability. This is defined as both a feature of the text, but also a capacity within a reader. So a text can be more or less followable, but in part this is a measure of a reader’s ability to stay with the account as it unfolds and then be able to respond to and deal with surprising twists and turns that their training or daily practices had not readied them for. Followability, as a feature of a text and also as a practice, this matter for scientific readers just as much as the literary.

Gallie also emphasises the importance of ordering for the purposes of telling a convincing story, and how later events need to be shown to have some extent depended on the earlier events, meaning that the earlier events were building a network of possibilities. Such an argument about the contingency of later events in relation to the past feels familiar to a line of argument developed by John Beatty in his contribution to the ‘Narrative in Science’ special issue, which focussed on evolutionary history. Beatty “Narrative possibility and narrative explanation,” Studies in History and Philosophy of Science 62 (2017): 31–41.

However, Gallie also at times tries to make a strict distinction between what is happening in literature and in science. This is clearest when it comes to the notion of a ‘conclusion’ which for storytelling, so he argues, is achieved in a starkly contrary way to conclusions in the sciences:

“The conclusion of any worthwhile story is not something that can be deduced or predicted, nor even something that can be seen at a later stage to have been theoretically or ideally predictable on the basis of what had been revealed at some earlier stage.”

On the face of it, this might seem very far removed from the purpose of a scientific conclusion, and indeed, Gallie was holding up scientific conclusions as a contrast class in order to create this definition. Nevertheless, we might want to consider whether scientific conclusions also share these characteristics. After all, plenty of experiments are performed without the aim of confirming a prediction. And even those that are pursued for the purpose of prediction confirmation often throw up results that are interesting and valuable in their own right, regardless of their relation to any set of predictions. Might it not be that literary narratives and scientific narratives simply sit on a continuum/spectrum when it comes to preferred style of conclusion? Perhaps it is a matter of convention or preferred narrative styles. The point here is not to answer definitively, but to point to the way such questions proliferate when we take a Narrative Science approach to these arguments. 

LINK TO PAPER

Nelson Goodman here builds on his earlier research in the philosophy of art, logic, and semiotics, to give an overall account of how the world or worlds we understand ourselves to be perceiving and engaging with, come to take the shape that they do, how our understandings of the world are structured or altered through the making of new representations, and as the title says, how worlds are made. Needless to say, this is a dense project, and even a short list of some of the key terms Goodman defines is enough to scare many people off: depictions; versions; visions; exemplifications; denoting; style; expression. Again, as with many of the philosophers surveyed here, Goodman makes a sharp distinction between what goes on in the sciences and everything else, on the grounds that descriptive verbal statements (as are the mainstay of the sciences) can have their truth or falsity assessed, whereas every non-scientific world making enterprise can only be said to be more-or-less right or wrong. But from the perspective of Narrative Science, any questions one might want to introduce about truth or falsity can be subordinated to the question of whether Goodman’s scheme, or at least some of his concepts, can help to better understand the signs, symbols, statements, and world versions that get put together in the sciences. 

LINK TO BOOK

A particular touchstone for historians and philosophers of biology, but addressing questions and presenting problems that matter for any historical enterprise, Gould’s Wonderful Life continues to motivate new research. It is here that he first published his argument-cum-thought experiment on ‘replaying life’s tape’, which asked the reader to consider the extent to which contingency influenced evolution and the forms of life that could come to emerge:

“I call this experiment “replaying life’s tape.” You press the rewind button and, making sure you thoroughly erase everything that actually happened, go back to any time and place in the past—say, to the seas of the Burgess Shale. Then let the tape run again and see if the repetition looks at all like the original. If each replay resembles life’s actual pathway, then we must conclude that what really happened pretty much had to occur. But suppose that the experimental versions all yield sensible results strikingly different from the actual history of life?”

The strength and depth of Gould’s commitment to contingency evidenced here is one that continues to provoke new responses, and deeper exploration of how the tape of life experiment works, and what else it might be teaching us. (See David Sepkoski  “Replaying Life’s Tape: Simulations, metaphors, and historicity in Stephen Jay Gould’s view of life,” Studies in History and Philosophy of Biological and Biomedical Sciences 58 (2016): 73–81, for an investigation of the origins and significance of the tape argument).

Beyond this particular example, much of Gould’s work is particularly friendly to the Narrative Science thesis, in part thanks to his dedication to the historical sciences, but beyond this, thanks to his integration of scientific, philosophical, historical and aesthetic forms of argumentation and knowledge-making. While Narrative Science is not the only kind of project to appreciate the latter, it is certainly well placed to develop it. 

LINK TO BOOK

One paper in a series that analyse the historical sciences, what they are, why they are distinctive, and how they work, Griesemer here makes a defense of their deeply theoretical nature. He does so by looking at the role of periodization in setting the terms of biological and evolutionary model making, demonstrating that periodizations are far from arbitrary or incidental with regard to the kinds of modelling or theory building that they allow. Moreover, inasmuch as practitioners might try out a range of different periodizations to find which best accounts for the phenomena, they can also be appreciated and evaluated as more or less robust.

In creating the terms for biological model- and mechanism-making then, Griesemer goes on to argue that these elements are incorporated into historical narratives of biological evolution and development. Taken together, periodization and narrative-making are shown to be deeply theoretical, helping to define the historical sciences:

“What is distinctive about historical science is that its periodization models coordinate narrative and mechanistic accounts and that historical scientists accept the narrative component as a contribution to understanding.”

In revisiting Griesemer’s account, we might want to further interrogate the relationship between mechanisms and narratives. Are we, for instance, using different words to describe elements of what is otherwise one thing?

LINK TO PAPER

 What are our methods for building historical and sociological accounts of contexts as they change over time or as we find them, and how do we assess their coherence and reliability? To get at these kinds of questions Hawthorn delves into counterfactual histories, argues for their usefulness, and goes on to present examples of counterfactual history making. The kinds of apparatus that counterfactual historians bring to bear on their cases reveal features and functions of narrative more broadly, all of which the Narrative Science project can pick up and adopt in investigations of science. How often, for instance, do scientists hold together many potential interpretations of a phenomena as plausible at any given time, and what guides their choice when they come to consider one to be more reliable than the rest? Are the same criteria used to assess plausible historical narratives also used in the sciences, or do scientists use different criteria, or are the criteria transformed in some way when moving to scientific contexts? These kinds of question can lead us to expand the range of ways in which we expect to find narrative knowledge present in science. 

LINK TO BOOK

Herman’s recent book serves as an excellent starting point for understanding narrative on the terms of literary theory and narratology. We are interested in the extent to which the features and functions of narrative, as identified and developed by those interested in literature, art, aesthetics, etc., can be fruitfully applied to scientific work and practices.

LINK TO BOOK

If one were looking for a Narrative Science sceptic, one need look no further than Carl Gustav Hempel. His account of historical and scientific explanation denied that narratives have the capacity to explain, on the grounds that a narrative only serves as a provisional ‘explanation sketch’, for a scientific placeholder for a proper explanation in non-narrative deductive-nomological statements. Moreover, his suggestion that narratives instead offer ‘understanding’, i.e. a way to grasp a complex situation about which we have only partial knowledge, was designed to create a sharper distinction between scientific and historical forms of enquiry. Both features of this philosophical position are brought into question by our thesis, though the former idea–that science only proceeds by improving and expanding the reach of deductive-nomological statements–has been dismissed for some time.

For a recent and in-depth historical account of the emergence of the Hempelian view, see FonsDewulf. “A Genealogy of Scientific Explanation: The Emergence of the Deductive-Nomological Model at the Intersection of German Historical and Scientific Philosophy.” PhD thesis, University of Gent, 2018.

LINK TO BOOK