5

Contemporary science presents knowledge as necessarily contingent rather than absolute, and increasingly context- and historically-specific.^[1] As such, the sciences promote the practice of continually critiquing, debating, and scrutinizing the current knowledge base through evaluations and assessments that (can) lead to modification, alteration and, at times, radical change in light of new evidence. The method(ology) of deciding what is “real,” “true,” “authentic,” and/or “accurate” in scientific practices remains contested, while the nature and reliability of human senses in scientific observations has been questioned since the ancient Greeks.^[2] Consequently, while scientific practice is embraced as a human endeavor, it is simultaneously framed as one that inevitably leads to personal prejudices, misapprehensions, and biases presented as inherent weaknesses, even vices, in rigorous scientific practice. These weaknesses and vices are seen as problems that must be overcome or mediated^[3] through various scientific practices and more recently, through implementing machine learning and automated systems.

Drawing on ethnographic fieldwork with astrophysicists (ranging from instrument scientists to astronomers and cosmologists) and microscopists in a United Kingdom university, I explore the everyday scientific practice of deciding “what is real.” I focus on everyday interactions between humans through and with technologies in open shared dialogue about the importance of “being human” and “getting it wrong.” I pose the question: do the sciences place too much faith in mediated visual images and/or computer simulations in deciding “what is real?” I highlight the key role of Aristotle’s practical wisdom (an intellectual virtue and inherently human endeavor) through Hans-Georg Gadamer’s concept of dialogue-as-play, a concept that extends beyond individual human bodies and requires an ability to listen and to embrace and manage deep uncertainty.

Setting the Scene: An Ethnographer in Astrophysics

I approach this topic as an ethnographer and human geographer working within the everyday world, which I understand to be complex, unfolding, and entangled (more than the sum of its parts).^[4] My interest is in what people do rather than what they think they do as they move through the everyday world.^[5] I focus on the minutiae, the detail, the seemingly mundane, and view the field of study to be historically and materially situated—that is, not confined by the four walls of the immediate environment. Ethnographers do not (initially) extrapolate our findings beyond our fieldwork setting. As such, descriptions of the field offer an important background against which to read our findings.

For eighteen months, I have been based at a United Kingdom academic institution, and specifically in a three-floor astrophysics building, designed by an award-winning international architect. The building houses communities of instrument scientists, astronomers/observers, and cosmologists from around the world.^[6] On the ground floor are the instrument research scientists, who are all trained astrophysicists, and a small number of space science engineers; the remaining engineers are based in an industrial site ten miles away. Many in this community work in the field of adaptive optics for large ground-based telescopes^[7] as part of multi-million Euro international projects. Some specialize in cutting-edge research into Free Space Optical Communications via satellites.^[8]

On the first floor are observers who mainly work in the field of extragalactic astronomy, rather than investigating our “nearby” solar system; they study stellar mass billions of light years away. Working on their desktop computers, the observers analyze and interpret data taken from astronomical instruments hosted at different observatories on mountaintop locations in locations like the Spanish islands, Chile, and Hawaii, and on space-based telescopes (for example, the Hubble). Each observer is trained in a highly specialized field of astrophysics, and studies a specific astronomical structure or phenomenon using data from a bespoke instrument that has been designed to capture targeted wavelengths at a specific point of the electromagnetic spectrum. Every year, a team or department of astronomers apply to space and/or ground-based observatories for time ‘on-sky’ to use a specific astronomical instrument for their research needs.^[9]

On the top floor are the cosmologists, mostly theoreticians and mathematicians, who develop computer simulations to test out theories of how (they believe) the universe works, and also to support observers in deciding where to point the telescopes to get the best results during a night of observing. Time on-sky at ground-based telescopes can be expensive, from €2,000 a night at smaller older observatories to €100,000 a night for larger modern observatories, with the potential for bad weather ruining astronomical seeing. Space-based telescopes provide higher resolution images, but to be allocated time here is highly competitive and costly. The cosmology simulations are calculated using supercomputers housed in their own three-story building, with a sophisticated water-cooling system on the roof mandated by the heat generated by the computers: one simulation run can take sixty-four computers running continually thirty-two days to complete. Many of the observers and cosmologists on the first and second floor work closely together within their own highly specialized fields. Younger generations of observers spend significantly less time on-sky (at observatories); increasingly their time is spent in front of computer screens, coding.^[10]

The building was designed around light, movement, and creating space for people to talk both informally (corridors, sofas, kitchens) and formally (small glass-fronted meeting rooms and break-out corners). The first and second floor have large open central spaces, with sofas and standing tables for coffee breaks and informal meetings. The offices are around the four walls, with sole occupancy reserved for permanent staff and shared offices for postdoctoral fellows and PhD students. Many people work with their doors open unless they have a meeting or don’t wish to be disturbed. The ground floor has no large open space and in its place is a long rectangular room, with glass on two sides, used almost daily (mostly by observers and cosmologists) for presentations or for postgraduate classes.

On each of the three floors, whiteboards are in every office and public space and are covered with drawings, sketches, and formulas. There is a wooden terrace with flower beds on the top floor; it is open to anyone in the building, although mostly the first and second floors use it. Every day at 11 a.m. and 4 p.m., the physicists on the first and second floors emerge from their offices and gather around the standing (poser) tables or sit on sofas. They exchange news, both social and work, and have coffee and biscuits together. Fifteen minutes before the morning coffee time, there is a short informal journal club to critique two or three of the many new papers uploaded daily to the international astronomy archive.

The ground floor has no designated coffee or meeting space, as the large meeting room takes up most of the central space. The claimed coffee space is a small open area beside the front door that was originally designed to welcome visitors. Notably, there is no whiteboard to sketch ideas on. Coffee times are ad hoc to fit in with lab times. The instrument research scientists interact less with the other floors, for a number of reasons including different working patterns. Core time is spent walking between their offices (where they create/craft computer simulations) and windowless laboratories in the adjacent building (to test the results of the simulations with on-bench models), before shipping instruments overseas to test on-sky (in astronomical observatories).

My office is on the ground floor; I share it with an astrophysicist/optical design engineer. Most days, I have “coffee and a chat” with the instrument scientists, and I usually walk/jog up the stairs to catch the fifteen-minute “astro-journal club.” I attend as many talks, journal clubs, informal discussions, or meetings throughout the building as I can. The meetings are many, and times overlap. Discussions are highly technical, and specialized and bespoke images, plots, and simulations are used as key discussion tools. Few images/plots have a standard format; many are colorful creations, and almost none have detailed labelling. A Q&A usually ensues to clarify the details.

As part of this fieldwork, I spent two weeks in an observatory (on-sky at night) in Spain with two instrument scientists (Cillian and Bob) who were commissioning (testing) an instrument as part of a Free Space Optical Communications study (see Figures 1 and 2). I also attended the largest biannual international conference on telescopes and space instrumentation (this year located in the United States). I participated in two weeklong workshops in the building: one with the adaptive optics community and one with cosmologists on the “realistic” nature of different mock galaxies (computer simulations of how the universe works) in large-scale surveys. During the first three months of the fieldwork, I spent time in a microscope suite and attended a three-day symposium in Germany with soft matter physicists, around the nature of microscope images. In this chapter, I draw mainly on findings from my time in astrophysics, beginning with a short scenario based on my fieldwork with Cillian and Bob^[11] (two instrument scientists/astrophysics) and the everyday question: “but is it real?”

“But Is It Real?”: Ethnographer’s Observations, Instrument Scientists’ Thoughts

Cillian has spent the morning walking back and forth between the two computer screens on his desk, where he is running simulation software, and the lab in the nearby building, where a camera is set up on a test bench. He is preparing to go on-sky to test the full instrument the following week in a British/Spanish/Dutch-owned mountain-based observatory in the Spanish islands. He did not buy and test the camera himself and so he is double-checking every aspect of the kit: shutter mode and type, frame rate as a function of size of image, and so forth. On the right computer screen he opens the black terminal screen to start a dialogue with the camera in the lab in the adjacent building. The terminal screen is empty, except for the blinking white cursor waiting for the first command. He pauses as he tries to remember the correct command. Cillian, like most of the instrument scientists and astrophysicists, prefers to work directly with the terminal. It shows all the inner workings of the software code. There are no algorithms built in to guide the user around the software, but working directly with basic commands and running code is considered more transparent and, therefore, makes it easier to find mistakes and faults.

While waiting for the camera (software) to respond, Cillian opens a power spectrum plot from research data taken at the observatory four months previously. He suddenly sits up straight, leans forward and stares at the power spectrum plot, and quietly says “the downlink is all noise.” Bob retorts “that’s interesting” without raising his head and carries on scribbling on a student’s draft paper in red pen. Cillian brings up another plot on the computer screen, and then another, and another—then says “it’s the same.” Bob carries on reading, scribbling and muttering under his breath. Cillian’s increasing silent frustration, and Bob’s non-response, means something is wrong. I move closer to look at the image. The plot has one horizontal line half way up the x-axis; it’s flat with no discernable pattern. Cillian brings up another plot on the screen, stares, sighs, puts his chin in his hand and says: “but is it real?”

Bob finally raises his head and leans over. They look at each other, look back at the plot on the screen, and Cillian says “…but why?” He opens the terminal on the right-hand screen and starts to “do calculations.” For ten minutes, he taps away, darting his head up and down from keyboard to screen, pausing briefly while calculations run, then continues tapping away. Bob comes over and asks, “have you tried…?” For the next thirty minutes, Cillian does numerous calculations on the computer, with Bob travelling the two meters to Cillian’s desk on his wheeled office chair with one well-honed push, then back to his desk to continue scribbling. Finally, Cillian sits back and says, “…it only goes up to 10 hertz….” Bob, without missing a beat, says “well that’ll be the problem.” They turn slowly to look at each other, then rock back in their chairs and chuckle (as I have seen many times on-sky, or in the office, when either something is going badly wrong, every attempt to solve a problem fails, or they have just solved a complicated problem). Cillian turns his head looks at me and laughs with relief.

To a passing visitor outside of the field, Cillian’s question (“but is it real?”) could be perceived as a short dialogue between Cillian and Bob, drawing on the plots, images, and calculations on the computer screen in that moment alone. Instead, this four-word question is part of the rigorous analysis of a six-year research project and provides a brief window into layers of work, discussions with others (in meetings, corridors, coffee rooms, lifts, via skype, slack, and emails); hours of experiments on-sky in astronomical observatories overseas; and moments grabbed reading thousands of short papers uploaded daily to the astronomy archive (accessed by astrophysics around the world and discussed daily at journal clubs over coffee).

Rejecting Authenticity: Constructed Nature of Images/Data?

When Cillian poses the question, “but is it real?,” he is not asking (himself), “is the plot authentic?” (That is, is the plot or image replicating or mirroring exactly what is “out there” in the external world and, therefore, offering up a perfect representation of the world for the viewer to grasp immediately.) As with microscopes, images gathered from all telescopes (ground or space-based) are constructed, created, and often enhanced through color, to create an image/tool that can only be interpreted by skilled observers through a shared forum as part of wider data analyses drawing on numerous sources of information. Few professional observers today “look” through a telescope and “see” objects in the night sky. Instead contemporary astronomers, or more accurately astrophysicists, observe through computer screens, either in the observatory through the hands-on skill and knowledge of a resident telescope operator and a resident night-time astronomer or, increasingly, in their offices hundreds or thousands of miles away.

Almost all contemporary astronomical images are constructed from information (light/photons) captured from targeted frequencies along the breadth of the electromagnetic spectrum; ranging from high-frequency, rarely-glimpsed gamma rays to slow-frequency, commonly-occurring radio waves (see Figure 3.) Each wavelength of photons offers different information and also (re)acts differently and, as such, bespoke specialized instruments are required to capture, and process the targeted photons. For example, collecting fast-moving gamma rays is akin to capturing a rare species of tiger that runs like lightning and moves round, over, and through every trap set in its path. It is difficult. If you don’t know where to observe, using a carefully crafted instrument, and you blink—it’s gone. In contrast, radio waves are more abundant. But while these wavelengths are easier to capture, they are also wider, and can only be collected by telescopes with large apertures the size of football fields.

Recently, in an attempt to create richer/more accurate data sources, images from different telescopes have been superimposed on each other. These images are referred to as multi-message observations, now a key funding term. For example, Figure 4, the composite image of the Crab Nebula, a supernova remnant, was created by combining data from five telescopes spanning nearly the entire breadth of the electromagnetic spectrum: the Karl G. Jansky Very Large Array (radio), the Spitzer Space Telescope (infrared), the Hubble Space Telescope (visible/optical), the XMM-Newton Observatory (X-ray), and the Chandra X-ray Observatory (Gamma). The color pallet is bright, human-made, and intended to create contrast between the different structures in the astronomical objects observed. The coloring of images is not new, nor is debate about the appropriateness of altering these images. The nineteenth-century astronomer T.H. Webb stated that his eighteenth-century predecessor William Herschel, the father of modern-day astronomy, was “rather too fond of red tints.” It is not known if Webb objected to the particular color used or to the act of changing the image at all.^[12]

Until recently, astronomers/”natural philosophers” observed within the visible part of the electromagnetic spectrum that forms only 10% of the spectrum as known today (see Figure 3).^[13] Theoretically, the human eye can perceive these wavelengths, but in reality, the raw data from captured photons is a blur of white light and grainy images (background noise) to the human eye. The raw data/images from optical telescopes require extensive post-processing in order to extract the signal from this background noise before any scientific interpretations can begin and meanings be inferred. Seeing/perceiving information does not equate to knowing^[14] or understanding.^[15] Moreover, for all ground-based optical telescopes, there is the added complication of atmospheric seeing. Looking up through the atmosphere, the images are distorted. To human eyes, the stars appear to move back and forth (known to astronomers as “wobble”) and the light intensity alters (“twinkle”). Adaptive optics corrects the former by reconstructing the image in the telescope, but currently there is no solution to “stopping the stars from twinkling.”^[16]

To identify, differentiate and assign meanings to colored telescope images is a skill in astronomical qualitative data (image) analysis, learned on the job and inaccessible to the uninitiated. There is a further perceived complexity to interpreting contemporary astronomical images, as compared to the astronomical worlds of Galileo, Newton and Herschel, who located and identified different types of structures in the “night sky.” These astronomers, who were also instrument scientists, cosmologists, and natural philosophers, lived before the carving-up of world knowledge into silos through academic disciplines.^[17] In marked contrast, contemporary professional astronomers and instrument research scientists are specialists guided by modern physics, driven by the desire to know how these processes are happening, and often view unmediated human vision as unreliable, unreplicable and untestible.^[18] As such, telescope images are considered to be only the first step in analysis. The image (qualitative data) is further transformed, often creatively, into various graphs, plots and diagrams (quantitative data) that are considered to be more transparent and reliable. As Cillian told me, after reviewing my fieldwork photos during an observatory visit and noticing I had photographed the telescope image and ignored the adjacent power spectrum plot:

When you first looked at [the computer screen in the observatory control panel] you saw an image and assumed that was the most important part … but that’s the least important part of the data … we have to try and process this data in order to decide if something is real or not or if it’s an artifact or which component fits into which … and then compare it to theory. If it doesn’t fit with theory then something is wrong.

Constructing Reality through Dialogue: The Good, the Bad and the Unknown

For these instrument research scientists in astrophysics, few images/plots are unquestioningly accepted as a trusted source of accurate or precise data. Instead a complex multi-layered dialogue begins by silently, often tacitly, asking the question: is this what we expected [based on our current understanding of what we believe to be true]? Rarely is the answer a resounding and unquestioning: yes, that is what we expected to observe. Here, the concept of “what is true” or “the truth” corresponds to current accepted theories in the field. But while these theories offer guidance during the dialogue as the instrument scientists decide if the image/plot is real, it would be a mistake for the reader to assume that these truths are doctrines, set in stone and immovable. When Cillian says “…if it doesn’t fit with theory then something is wrong,” he is not implying that the theory is inherently correct and the image is incorrect.^[19]

To begin deciding if sources of information (image, plot, diagram, conversation, or material entity) are real, Cillian and others in the astrophysics community engage in the analytical and methodological practice of revisiting what has happened before, usually starting by identifying sites of potential mistakes and known systematic and systemic errors. What triggers this process is noticing that something is “not quite right,” and the immediate reaction is often visceral: a sudden straightening of the back, drawing back of shoulders, peering into a computer screen, and/or a long deep sigh. But before they can conclude whether or not “it is real” and before the overarching dialogue begins between what has been observed (in the data) and what is expected to be observed (in the theory), a further (fundamental) question needs to be answered: is the data in the image/plot accurate/precise? For Cillian and Bob, the chosen term depends on whether they are building and testing the bespoke instrument (low technology readiness or LTR) or testing the capacity of the final instrument (high technology readiness or HTR). During LTR, measurements must be precise, so that they know each component of the instrument, no matter how small, is doing exactly what they intended it to do. Through testing every working part, from simulations to on-bench and finally on-sky, the data collected at every step must be precise, although not necessarily accurate. It can even be wrong (inaccurate). The term accuracy becomes important during the HTR stages and for observers when analyzing the data from an on-sky instrument.

When Cillian and Bob began checking the precision of the data in the power spectrum plot, as described earlier, every step that had created the plot was traced back and forth and discussed in minute detail; every calculation was checked and double-checked and became part of a complex multi-layered dialogue. They drew on their own knowledge and experience (for example, conversations with their past selves through lab diaries and individual or shared memories) and that of other colleagues (archived web-based discussion threads and diaries) across time and across space (for example, a telephone call or real-time virtual discussion threads with the international community). The instrument scientists that I have met will pick up a phone without hesitation to call someone in another country. Often they “ping a message” through a real-time chatline (including slack or google hangouts) or “fire off” an email. Responses back are often immediate. These networks of communication are informal, fluid, highly complex, and multiple in nature.^[20] Dialogues stop and start spontaneously due, in part, to the work schedule of many instrument scientists (and engineers), who work on multiple projects at the same time and have to learn how to interweave projects together, grabbing moments to complete different tasks.

I have observed similar multi-layered dialogues in meetings or talks with observers and cosmologists, either together or independently, where the focus is on the accuracy of the data, the sensitivity of the instruments, and the selection bias of samples. During the week-long workshop on mock galaxies, the cosmologists discussed the reality of theoretical simulation, some aspects of which cannot be ratified by observational data as the instrumentation has not yet been developed to collect the faint light that has travelled from distant astronomical structures. Instead, discussions revolve around the reality of the images in relation to theoretical knowns and unknowns. Intense debates can break out over whether the error is systematic or systemic—a necessary distinction so that the cosmologists know how to account for these errors in the final analysis. Underpinning many discussions is a deep concern that the simulations should represent the observational data. During presentations on new ways of modelling the universe, if a presenter fails to highlight any errors or unknowns, it is not uncommon for seasoned cosmologists to ask, rhetorically: “Well, your model is elegant, but can it cope with the messiness and chaos of the real universe?”^[21]

Similar multi-layered dialogues occurred in the microscope labs. Jen is a professional microscopist with over thirty-five years’ experience, particularly on the electron microscope. During extensive analysis sessions, she would lean forward, peer at the image in the microscope (shown on a 20-inch computer screen) and ask (herself): “uhm … is it real?” In these situations, she has noticed something unexpected and is trying to determine if this unusual object in the image is a foreign body (an artifact), and has been introduced during the preparation process, or is part of the original biological specimen. In one situation, Jen shouts to the assistant in the next room: “who stained the sample, you or a student?” When the return answer was “me,” she mutters: “okay, so no contamination there…(long pause)…Did we change the people supplying the grids? The last lot were no good. We did? Okay, so it’s not that.” The assistant, with whom she has worked for ten years, walks into the microscope room. They both peer at the computer screen and carefully consider each potential site of contamination during the preparation process that has taken Jen over a week to complete: how the sample was prepared, were the fluids fresh. Every minute detail of how the final sample had been prepared, how it was placed in the tube and then in the microscope in that specific moment and time, was questioned—until they decided together that the foreign body (artifact) was part of the original biological sample. Watching these dialogues unfold was similar to observing detectives at work, looking over the shoulders of Miss Marple, Hercule Poirot, or Sherlock Holmes. Not all dialogues played out like this. For example, I observed a visiting PhD student silently panic when Jen began asking routine questions about the origin of the biological sample as she started to prepare the sample for the microscope. The student looked away into the corner and responded: “My supervisor thinks everything is okay, they just want me to run these…. I just need one more set of results…supporting previous findings.”^[22]

These descriptions of the style of dialogue may indicate a linear, rational, and cognitive process for deciding what is real. In practice, however, these dialogues are complex, fluid, highly intuitive, iterative, and embodied. Rather than thinking-then-doing, fieldwork participants (instrument scientists, astrophysicists, and microscopists) were frequently thinking-through-doing.

Thinking-Through-Doing: Whiteboards, Dancing, and Origami

Throughout my time in the field, the physicists that I spent time with would often begin explaining an idea to me, then would pause, pick up a whiteboard marker, and ask “may I?” I would nod, and they would start to draw, sketching out ideas considered too difficult to describe verbally. Or, as I observed, they were often thinking through their ideas as they drew/sketched. Working publicly on a whiteboard can be difficult for new students as they have to share their work openly. Sometimes I would meet students sketching on whiteboards in open spaces late in the evening so that no one would see their work, and they always rubbed it off before leaving.

At times, I found myself completely outside a whiteboard conversation (or whiteboard party, as Cillian calls them) not because I lacked knowledge of a specific concept or I couldn’t read the mathematical formulas, but because the astrophysicists (cosmologists, observers and instrument scientists) would communicate in half-sentences. One speaker would begin a sentence and after a few words, the listener would say “‘oh yes”’ or “you mean, like this,” and draw a sketch.   One memorable moment occurred when an observer and a cosmologist, who work closely together, solved a thorny problem through a high-speed dialogue composed of words, hand and body movements, and silent mutual staring at the board, interspersed with sketches and formulas that covered a wall-sized whiteboard in a matter of minutes. In remembering thise episode, I am reminded of watching a video of the abstract artist Jackson Pollock at work. Pollock covered expansive canvases in streaks of seemingly random brushstrokes and splashes of paint, in no particular linear order, working close to the canvas, but not stopping in one position for long. When he finished and the camera zoomed back, the work, to me, formed a harmonious whole. At the time, while watching the observer and cosmologist, I was fixed to the spot, mesmerized, in particular, as their individual bodies, speech, and hands appeared to merge into one being.

I have encountered similar experiences of being with Cillian and Bob in the mountain-top observatories when they were commissioning new instruments. They work between two observatories, located 426 meters apart, and “jump in” a car to make the five-minute journey. Once at the observatory, they run up and down stairs and through a maze of rooms performing tasks (for example, searching for or testing, tracing, or installing cables) that require a keen understanding of the instrument, the observatories, and the research project. If additional knowledge/experience is needed to solve a problem, no matter how small, they “call a friend” in the community. Moving at high speed, they communicate almost telepathically. At times they left me scrabbling to gather my things (camera, bag, coat) as, after pondering a problem together, they suddenly got up and sprinted out the door.^[23] I have numerous photos of them working together. At times, it is difficult to tell which hands or arms belong to which body; the biological and the technological merge as they work out the problem (see Figure 5). Afterwards, when we review fieldwork photos, they have limited memory of what has happened.^[24]

Whiteboard parties are not the only way that (astro)physicists  think through ideas. One of the cosmologists in the building uses the Japanese art of paper folding (origami) to explain the nature of the cosmic web.^[25] Commentators compare his use of origami to an analogy, a scientific show and tell, to explain to others the interconnected foldings in different galaxies. But when we spoke together about how he developed this concept, it emerged that endless hours of paper-folding (he taught himself from online videos), and experimenting/playing with different patterns of folding, had enabled him to think through the layered nature of the cosmic web. This embodied way of being/thinking mirrors the work^[26] of the foundational astronomer William Herschel, who was also a professional musician, composer, and teacher.^[27] It is thought that Herschel composed symphonies as a way of exploring his thinking on the complex and unknown nature of the universe. Herschel’s younger sister, Caroline was an accomplished astronomer and singer, and the twentieth century physicists Max Planx and Albert Einstein were outstanding musicians who extolled the virtues of music in their working lives. Perhaps unsurprisingly, many of the physicists in the building are also highly accomplished musicians, singers, photographers, and dancers, some to professional standards. When explaining concepts in talks or over coffee, they frequently use their hands, arms, or upper or whole bodies to imitate the dynamic nature of different astronomical structures and processes. Many engage in a form of “interpretive dance.” However, precision in choice of words is also a prized skill, and woe betide any student who uses the wrong term in a talk. Words matter, and playing with words and creating new ones is part of their everyday practice, in particular in devising acronyms for new research projects and instruments. But words (spoken/written) are only part of the dialogue.

I have also observed this embodied way of working with experienced microscopists, who, when observing and analyzing images, focus and refocus the microscope by “twiddling” the fine focus dials, simultaneously scanning back and forth across the grid. This requires extreme dexterity and they are often working with both hands at once, each in different directions, as difficult as the classic exercise of patting the head and stroking the stomach. This is a form of thinking-through-doing, an important ritual recognised by (some) microscope designers, who in redesigning a high-resolution full-automated microscope choose to keep the manual fine-focus dial to allow the skilled microscopists to make the final adjustments to the image. As a designer and former microscopist told me: “it [the manual fine-focusing] is part of observing, analyzing the image.”

This concept of thinking-through-doing with and through the material world has been well documented, in particular in the arts.^[28] The anthropologist Tim Ingold has written about “thinking-through-making,”^[29] although he extends this concept only to his definition of “makers” (artists, architects, and craftspeople) and openly dismisses the sciences.^[30] In contrast, I propose that this concept can be widened and better explicated through the work of Hans-Georg Gadamer, an eminent twentieth century German hermeneutic phenomenologist, and his lesser-known concept of dialogue-as-play. For me, this dynamic concept and form of practical wisdom illuminates the nature of many rigorous dialogues around deciding what is real in everyday scientific practices, as encountered in the field.

Dialogue-As-Play: Everyday Practical Wisdom Beyond Spoken Words

Hans-Georg Gadamer (1900–2002) was a leading continental philosopher. His work in “philosophical hermeneutics” extended our understanding of our everyday lives and explored the detail of everyday existence through shared dialogue. Gadamer’s dialogue is dynamic, dialectic, and interpretive. Key elements of authentic dialogue include openness toward others and respect for differing viewpoints, gained through listening.^[31] Crucial ethical conditions of genuine understanding include being open to learning new ideas from others that could radically change our understanding of the world around us.^[32] Gadamer employed the analogy of play, in which the “players” are caught up with, and lose themselves in, their experiences, a dynamic that holds the potential to transform all players. This form of dialogue moves beyond words as instrumental and engages all our human senses, beyond the contemporary five.^[33] These dialogues are not confined to human beings alone and engage the material world (including the visceral body and everyday tools/technologies).

As we move through and interact with our everyday worlds, we travel back and forth through (historical) time and space, encountering embedded rituals, traditions, prejudices. Some we subvert through devising “go-arounds,” others we embrace or choose not to or cannot subvert/alter for whatever reasons. Dialogue-as-play is dynamic, never-ending, often based on tacit knowledge, and therefore remains partially or wholly unspoken. Here, the concept of play is not akin to the contemporary view of (child’s) play: “to engage in activity for enjoyment and recreation rather than a serious or practical purpose.”^[34] Importantly, the playing-of-the-game rather than the player alone co-creates meaning through shared dialogue^[35] that, following Aristotle, privileges listening above the seeing/speaking that has dominated Western philosophy.^[36] Underpinning dialogue-as-play is the practice of approaching dialogue knowing that one does not and indeed cannot know everything.

Following Aristotle, Gadamer proposed that all philosophy starts from praxis (human practice) and that hermeneutics is essentially practical philosophy. In its simplest form, according to Aristotle, practical wisdom is the ability to negotiate and make sense of the everyday world through thoughtful deliberation, judging well for the common good, and acting decisively with foresight.^[37] For Gadamer, human beings must continually interpret their everyday world, as we are neither neutral, independent, nor objective observers, but rather existential finite interpreters, always expressing linguistically our relation to the world.^[38] Gadamer is often perceived to have focused on language as only spoken words, which he preferred over written text.^[39] In a recent paper on the nature of dialogue in the “good” medical consultation, I extend dialogue-as-play to include the somatic body and technologies-to-hand that, I argue, are not at odds^[40] with Gadamer’s work.^[41]

During this fieldwork, however, I observed intense disagreement during high-energy discussions that, on first appearance, seems to clash with Gadamer’s work, as he viewed dialogue-as-play to be about solidarity^[42] and partisan ways of being.^[43] The French philosopher Jacques Derrida, one of Gadamer’s fiercest critiques, lambasted him for promoting the idea that there is one truth and that through dialogue there can be consensus and agreement.^[44] Derrida argued that there remained, always, a violence expressed through diversity of viewpoints.^[45] However, Gadamer’s dialogue-as-play within a “fusion of horizons” leads to the merging (not augmenting or cohering) of different worldviews through dynamic modes of interpretation that create new ways of understanding. The core to this concept is agreement on the matter under discussion, but this can translate into agreeing to come together to engage with the topic openly while listening to the other(s), including agreeing to disagree.^[46]

Diversity of views, and creating space for them, is the epitome of good scientific practice. Dialogue-as-play is needed to negotiate the uncertainties in scientific research in the everyday world, by being certain about what we do and do not know.^[47] Moreover, in contrast to microscopists or biologists, experimental astrophysicists and instrument scientists cannot set up or carry out traditional lab experiments, nor can they (arguably) control their sample or any variables. The astronomical observatories atop mountains are, in effect, their labs, and variables cannot be perturbed or controlled within the vastness of the night sky (or the day sky, for solar telescopes). Within this dynamic environment, where conditions cannot be replicated, the calibration process and knowing the precision, accuracy, and sensitivity of the instrumentation is paramount.

Practical Wisdom thru Dialogue-as-Play: the Unknown, Transparency, and Design Values

Unknowns and known unknowns dominate, and are openly discussed, within the specialized fields and scientific practices of astrophysics. This practice can be extended to microscopy and, arguably, to the diverse fields that make up “the sciences”. Engaging in rigorous everyday scientific practice is complex, diverse, and challenging; therein lies its beauty. In response to the unknown, unexpected, and unsolved, and key to rigorous analyses, experienced instrument makers/astrophysicists in this fieldwork, unapologetically embrace discuss and (attempt to) account for primary mistakes (in contrast to stereotypical portrayals in the media). This practice leads to new concepts and ideas, from which new directions can evolve. As such—and here I return to where I began—contemporary “good” scientific practice views knowledge as inherently contingent rather than absolute in nature. Consequently, within everyday scientific practices the skilled craft of engaging in dialogue-as-play is considered important in deciphering which sources of information are real i.e. data, images, papers, informal and formal discussions. (Here, dialogue moves beyond individual bodies, spoken words, and gestures to include the somatic body and technologies-to-hand.) Through dialogue-as-play, we make sense of, engage with, submit to, accept, challenge, and alter (although never “tame”) our dynamic, chaotic everyday world. Importantly, through dialogue-as-play scientists (can) gain, maintain and develop practical wisdom, leading to “good” or virtuous everyday scientific practices.

Methodologically, practical wisdom is challenging to study empirically, as it is enacted in the moment, cannot be taught, and is fluid and temporary in nature and, therefore, often remains unspoken. Initial findings indicate that potential conduits for practical wisdom (in this field) include: creating space(s) for rigorous questioning of the “reality” of each information source; an informal mentoring system that counteracts the normative concept of “the correct answer” and encourages individuals/groups to face any fear of failure, to speak out, put ideas forward, and engage critically; and developing transparent and open technical systems, accessible to an interconnected network of makers and users, as required.

Inevitably, there are barriers to dialogue-as-play, some of which are (deeply) embedded within our everyday lives in some form or other. For example, mainstream science education,^[48] policy documents, public engagement events, and the media frequently, albeit unwittingly, often present the sciences as a monotheistic set of disciplines, based on various forms of absolute knowledge, with one common language. Here, dialogue-as-play rarely thrives since dialogue appears less about listening openly, understanding the other, and accepting that one cannot know everything, and is more about convincing non-believers of the one truth. Any mentoring system also need to be bespoke to the learner as few methods fit all learning styles, for example, the “sink or swim” model. Returning briefly to the question I raised at the start of this chapter, one embedded barrier to dialogue-as-play is the practice of assigning too much (blind) faith to the meaning of an image, plot, or graph, without critical detailed analysis within an interconnected network of information sources. Notably, however, intense critical questioning, outlined in the earlier scenario with Cillian and Bob, was a common occurrence in this fieldwork (with the exception of inexperienced students). But for reasons beyond the scope of this chapter, there is an increasing divide and decreasing communication between the makers of instrumentation and the observers who use the kit and who increasingly work away from the observatories.^[49]

Underpinning these barriers is the omnipotent presence of an oxymoron in the sciences, mentioned at the start of this chapter: scientific practices are framed as a human endeavor, but one that views human senses and sensibilities as inherent weaknesses, even vices, in rigorous scientific practice. These weaknesses and vices must be overcome or mediated through various scientific practices, including machine learning and automated systems. Consequently, future questions include:

1. If scientific practices are inherently a human endeavour, then how do we learn to trust and reintegrate human senses into everyday scientific practices that are considered rigorous, reliable, and ethical?
2. In an increasingly specialised and fragmented world of knowledges, disciplines, and subdisciplines, how do we decrease the widening division between those who design, build, and develop and those who commission and use?
3. If practical wisdom cannot be taught, and must be learned on-the-job and enacted in the moment, how do we “teach” machine learning systems that currently rely on well defined, logical, generic algorithms?

FIONAGH THOMSON is an ethnographer interested in what people do rather than what they think they do. A former research associate at the University of Notre Dame, her background is in visual anthropology, educational philosophy, environmental and medical ethics, and human geography. Thomson’s projects are transdisciplinary and her fieldwork includes a variety “everyday spaces” ranging from the Scottish islands, the corridors of NHS hospitals, laboratories, and patients’ homes to rural woodlands, urban streets, and mountaintop astronomical observatories. Fionagh is currently a visiting researcher at the Centre for Advanced Instrumentation, Fundamental Physics, Durham University, UK.