Over the past several centuries, quantification has become one of the most powerful tools for research on water, just as it has for many other areas of science. Some of the power of quantification comes from its demonstrated practical utility; some of it comes from its speculative potential. In a variety of political and cultural contexts, the precision of numerical description has promised to effect an epistemological phase change, transforming an unruly, locally variable set of fluids—“waters” in the plural—into a singular substance—“water”—that can be weighed, measured, and classified with confidence.1→Christopher Hamlin, “‘Waters’ or ‘Water’? Master Narratives in Water History and Their Implications for Contemporary Water Policy,” Water Policy 2, no. 4–5 (2000): 313–325.
→Jamie Linton, What Is Water? The History of a Modern Abstraction (Vancouver: UBC Press, 2010).
→Jessica Barnes and Samer Alatout, “Water Worlds: Introduction to the Special Issue,” Social Studies of Science 42, no. 4 (2012): 483–488.

It is largely because of this quantification and singularization of water that it is possible to speak meaningfully of a multidisciplinary domain of “the water sciences,” as opposed to a kaleidoscopic set of largely incommensurable sciences of waters. But water’s multiplicity has not disappeared; it has simply gone underground. As hydrologists have long been trying to teach us, there is no way to account for the water we see without tracing its links to the waters we don’t. This brief essay explores the fluid boundaries between quantitative and qualitative understandings of water through the work of one of those hydrologists, Luna Leopold, who was a leader of the turn to quantitative methods in the study of rivers in the United States in 1950s and 1960s.

The quantitative turn in the study of rivers

“At perhaps no moment in the history of the systematic study of water have scientists been more enamored of quantification, as a path to certainty and authority, than they were in the middle decades of the twentieth century.”

The period of the 1950s and 1960s was a watershed in the history of the study of rivers in the United States and indeed throughout the world. At perhaps no moment in the history of the systematic study of water have scientists been more enamored of quantification, as a path to certainty and authority, than they were in the middle decades of the twentieth century. Laboring in the shadow of physics—widely seen at the time as the most scientific of the sciences—hydrologists sought to translate any and all scientific claims about water into quantitative, mathematical, and physical terms, while shunning as “unscientific” any claims that could not be so translated. One of the culminations of this ambition was the UNESCO-led International Hydrological Decade of 1965–1974, which helped to consolidate hydrology throughout the developed and developing worlds as a fundamentally “physics-like” discipline—that is, one that was quantitative, mathematical, and mechanistic.2Raymond L. Nace, Water and Man: A World View (Paris: Unesco, 1969).

This was neither the first nor the last time that water scientists have sought to distinguish themselves by outnumbering their predecessors. Another such moment, perhaps the most important, spanned the late eighteenth to the early nineteenth century, when water was first identified as the compound H2O. This new nomenclature and new understanding of matter gave scientists a way of describing multiple, incommensurable “waters” as variations on a single chemical substance, to which standardized metrics could be applied.3→Paul Needham, “The Discovery That Water Is H2O,” International Studies in the Philosophy of Science 16, no. 3 (October 2002): 205–226.
→Hasok Chang, Is Water H2O? Evidence, Pluralism and Realism (Dordrecht: Springer Verlag, 2012).
Nonetheless, the period spanning roughly from the 1930s to the 1960s represents a notable peak in this longer history of quantitative enthusiasm. This was a moment in which quantification appeared to be not merely one of the best ways to produce scientific knowledge about water, but, indeed, the only way.

“Before the 1950s, with only a few exceptions, most studies of the forms of natural rivers had been qualitative, focusing on the sequence of geological and climatic shifts responsible for a river’s observable present-day features.”

The turn to quantification affected even those areas of the science of hydrology that might seem least amenable to it, such as the study of processes that shape river channels, river valleys, and floodplains. This subject, fluvial geomorphology, had long seemed resistant to the kinds of quantification long applied to human-engineered canals and irrigation channels because of the complexity of the climatic and geological factors that determined the behavior of any given river in its particular place and time. Before the 1950s, with only a few exceptions, most studies of the forms of natural rivers had been qualitative, focusing on the sequence of geological and climatic shifts responsible for a river’s observable present-day features. Fully “scientific” by the standards of their time, such studies came to be seen by the hydrological and geomorphological quantifiers of the mid-twentieth century as quaint and outdated.4T.P. Burt et al., eds., The History of the Study of Landforms or the Development of Geomorphology, Vol. 4, Quaternary and Recent Processes and Forms (1890–1965) and the Mid-Century Revolution (London: Geological Society of London, 2008). Today’s fluvial geomorphology continues to be deeply shaped by the consequences of this quantitative turn.5Ellen Wohl, “Time and the Rivers Flowing: Fluvial Geomorphology since 1960,” Geomorphology 216 (2014): 263–282.

The limits of quantification

For those mid-twentieth-century quantifiers, the critical mistake of an older form of fluvial geomorphology was precisely its assumption that the climatic and geological setting of a river was critically important. On the contrary, they argued, almost everything that was important about a river could be determined by focusing on the dynamic relations between variables, such as the depth and width of the river channel and the rate at which water and sediment flowed through it. Those were just the kinds of variables that could be precisely measured in the field and in the laboratory, and it was that task of making precise measurements to which the rising postwar generation of fluvial geomorphologists devoted themselves.

https://items.ssrc.org/wp-content/uploads/2020/10/Luna_Leopold-scaled.jpg
Luna Leopold in the mid-1970s along the East Fork River in Wyoming. Photo: UC Berkeley/US Geological Survey.

Leopold was one of the leaders of that generation; as coauthor of the influential textbook Fluvial Processes in Geomorphology and head of the most well-funded and prestigious US institution for water research in the 1950s and 1960s, the USGS Water Resources Division, he shaped the field as a whole.6Luna B. Leopold, M. Gordon Wolman, and John P. Miller, Fluvial Processes in Geomorphology (New York: Dover, 1964). His fellow geomorphologist, Robert Sharp, observed that Leopold was “always advocating the measurement of something”—a description that could have applied just as well to a number of his contemporaries.7Robert P. Sharp to Penrose Medal Committee, Geological Society of America, 3 December 1991, box 9, folder 15, Papers of Robert P. Sharp, California Institute of Technology.

With the benefit of hindsight, there are multiple ways to criticize the almost monomaniacal focus on numbers of Leopold and his contemporaries, even while acknowledging that their methods remain a very productive way of studying rivers. One way is simply to note its incompleteness. Even at the time, and without transgressing the well-policed boundaries of scientific discourse, there were those who argued that the pendulum had swung too far toward the quantitative, putting other hard-won bodies of knowledge and skill at risk. The geologist J. Hoover Mackin, for example, posited that the “swing to the quantitative” in fluvial geomorphology—which he thought was epitomized by the work of Leopold and his collaborators—risked overshadowing the more essential work of reasoning carefully about cause-and-effect relationships in nature, even when those relationships could not be easily quantified.8J. Hoover. Mackin, “Rational and Empirical Methods of Investigation in Geology,” in The Fabric of Geology, ed. Claude C. Albritton, Jr. (Stanford, CA: Freeman, Cooper and Co., 1963), 135–163. Restricting “science” to the collection of quantitative data, he argued, drew attention away from critical problems, such as the development of landscapes and geological strata over long stretches of the Earth’s history, that were not amenable to quantification with the tools then available.

“As the momentum of the environmental movement grew over the course of the 1960s, the idea that everything that mattered about rivers could be summed up in numbers came under increasingly vocal attack.”

Outside the boundaries of science, the incompleteness of quantitative fluvial geomorphology came into even starker relief. As the momentum of the environmental movement grew over the course of the 1960s, the idea that everything that mattered about rivers could be summed up in numbers came under increasingly vocal attack. Rivers, environmentalists argued, were not merely physical phenomena produced by mechanistic interactions between water and land; they were also beautiful and vital components of life-worlds that included humans and other creatures. Even among the arch-quantifiers, this was a viewpoint that had some purchase. Leopold, for example, argued passionately over the course of his career that certain things about rivers, indeed some of the most important things about rivers, could never be captured by numbers, and certainly could never be expressed in terms of dollar values. Other “values”—aesthetic and ethical ones, which were either impossible to quantify or, if quantifiable, then impossible to compare meaningfully to economic values—would also have to be taken into account.9→Luna B. Leopold, Water and the Conservation Movement (Washington, DC: US Geological Survey Circular, 1958).
→Luna B. Leopold, A View of the River (Cambridge, MA: Harvard University Press, 1994).

Beyond the qualitative/quantitative divide

Given the imperialistic tendencies of reductionist sciences—that is, their tendency to assert that methods that have worked in certain contexts will work everywhere—these kinds of critiques of the limits of quantification are well worth making. They make it clear that the universal scope of quantitative science is a project, not a premise, and a highly incomplete (and almost certainly incompletable) project at that. In some ways, however, the critique only reinforces the very tendencies it is trying to challenge: It divides the production of knowledge into two camps or schools, the quantitative and the qualitative, and claims that each has its own unique characteristics. Once this division has been established, it is all too easy for the partisans of each side to claim that theirs is the only path to true understanding. In the debates that follow, the quantifiers somehow always seem to win. Indeed, as the example of Leopold proves, the quantifiers can argue from both sides, ensuring that even as certain unquantifiable values are protected, other values are quantified (and exploited) to the fullest.

But there is another way to think about the limits of quantification in the water sciences, and in science in general, that neither runs afoul of the same structural tendency toward dichotomy and hierarchy, nor seeks solely to preserve an enclave for nonquantifiable values in a world fundamentally transformed by quantification. Namely, one can take note of the fact that the foundations of quantitative science themselves are profoundly and, one might argue, inevitably nonquantitative. Numbers themselves are not the dry, disembodied abstractions that scientists committed to a certain model of epistemic and social authority would have us believe them to be; nor are they limited in precisely the ways that partisans of qualitative understandings of nature argue they are. On the contrary, quantitative and mechanistic understandings of the natural world emerge, like any other kind of understanding, from scientists’ embodied encounters with that world.10Durham, NC: Duke University Press, 2015More Info → Moreover, the numbers that are generated in such embodied encounters themselves have the capacity to become lively, affect-laden participants in that world.11Kristoffer Whitney and Melanie A. Kiechle, “Introduction: Counting on Nature,” Science as Culture 26, no. 1 (2017): 1–10.

Take Leopold, for example, who argued as passionately for the collection of quantitative data on rivers as he did for the preservation of those very same rivers’ intangible ethical and aesthetic values. In that sense, he was a good son of his father, Aldo Leopold, who first became influential among conservationists for analyzing wildlife as a “crop” to be rationally harvested and later was heralded by US environmentalists for his lyrical appreciations of the “integrity, stability, and beauty” of wild nature.12→Aldo Leopold, A Sand County Almanac, and Sketches Here and There (New York: Oxford University Press, 1949).
→Julianne Lutz Newton, Aldo Leopold’s Odyssey (Washington: Island Press/Shearwater Books, 2006).
In the actual practice of his science, Luna Leopold was anything but the disinterested number-cruncher that his identity as a leading advocate of the quantitative revolution in fluvial geomorphology might suggest. On the contrary, as the meticulously constructed personal diaries that he donated to the archives of the American Philosophical Society reveal, he was someone who reveled in the opportunity to immerse himself in the rivers he studied.

“His immersion in data was more than metaphorical; it was a way of life linked to a particular vision of the land and of society.”

In those diaries, hand-bound in leather and filled with photographs, sketches, poems, recipes, and even a few pressed leaves and flowers, as well as detailed descriptions of field expeditions and travels abroad, one gets a fuller picture of Leopold’s scientific practice than what can be found in the dry scientese of most of his published work. Mid-twentieth-century philosophers of science made much of the distinction between the “context of discovery” and the “context of justification,” arguing that while scientists’ hypotheses about nature might have all sorts of untidy, irrational origins, the process of determining which of those hypotheses accurately described nature was orderly and rational.13→Hans Reichenbach, Experience and Prediction: An Analysis of the Foundations and the Structure of Knowledge (Chicago: University of Chicago Press, 1938).
→Naomi Oreskes, “A Context of Motivation: US Navy Oceanographic Research and the Discovery of Sea-Floor Hydrothermal Vents,” Social Studies of Science 33, no. 5 (2003): 697–742.
In Leopold’s diaries, that neat and useful distinction falls apart, as it becomes clear that he returned to the field over and over again to take certain kinds of quantitative measurements precisely because of his love for particular rivers, for particular kinds of embodied experiences, and for the people with whom he shared those experiences. In this sense, his immersion in data was more than metaphorical; it was a way of life linked to a particular vision of the land and of society.

Living with multiplicity

The embodied, situated, and affective aspects of the kinds of quantitative scientific work carried out by Leopold and his collaborators do not render the limitations and blind spots of that kind of research any less important, nor do they wash away the deeply problematic commitments to economic development (albeit constrained) and scientific expertise (albeit qualified) that shaped most such work in Leopold’s time. By acknowledging them, however, we can see something that is obscured when we accept uncritically the idea of a divide between quantitative and qualitative ways of knowing water, or of knowing the world. Namely, we can see that the oppositions that really matter are not those between science and beauty, facts and values, disinterested knowledge and ethical commitment, or objective observations and subjective impressions, but rather between whole packages of beautiful (or ugly) science, valued (or disvalued) facts, knowledgeable (or ignorant) ethics, and sensitive (or insensitive) encounters. Quantitative claims about water, like any other claims, are “discovered” in the course of embodied, affective encounters with the world and “justified” through processes that are no less entangled with that world.

In this way, we come full circle to the idea of water’s multiplicity. Water is multiple not just because it is physically protean and interpretively flexible, though of course it is both. It is multiple because every way we have of knowing water—including the most rigorously quantitative, mathematical, and mechanistic ways imaginable—is inevitably and inextricably bound to how we value water, which in turn is shaped by our embodied encounters with water and with other beings. In other words, we do not live in a world where different values and interpretations can be applied to the same physical substance, H2O, but in a world replete with a kaleidoscopic variety of bundles of embodied affects that share just enough of a family resemblance that we can agree, for convenience’s sake, to call them all “water.” In the former kind of world, we might imagine that disagreements over water could be resolved when all parties agree on a set of basic facts and principles of reason; in the latter kind of world, the world we actually live in, there are no such easy solutions.

References:

1
→Christopher Hamlin, “‘Waters’ or ‘Water’? Master Narratives in Water History and Their Implications for Contemporary Water Policy,” Water Policy 2, no. 4–5 (2000): 313–325.
→Jamie Linton, What Is Water? The History of a Modern Abstraction (Vancouver: UBC Press, 2010).
→Jessica Barnes and Samer Alatout, “Water Worlds: Introduction to the Special Issue,” Social Studies of Science 42, no. 4 (2012): 483–488.
2
Raymond L. Nace, Water and Man: A World View (Paris: Unesco, 1969).
3
→Paul Needham, “The Discovery That Water Is H2O,” International Studies in the Philosophy of Science 16, no. 3 (October 2002): 205–226.
→Hasok Chang, Is Water H2O? Evidence, Pluralism and Realism (Dordrecht: Springer Verlag, 2012).
4
T.P. Burt et al., eds., The History of the Study of Landforms or the Development of Geomorphology, Vol. 4, Quaternary and Recent Processes and Forms (1890–1965) and the Mid-Century Revolution (London: Geological Society of London, 2008).
5
Ellen Wohl, “Time and the Rivers Flowing: Fluvial Geomorphology since 1960,” Geomorphology 216 (2014): 263–282.
6
Luna B. Leopold, M. Gordon Wolman, and John P. Miller, Fluvial Processes in Geomorphology (New York: Dover, 1964).
7
Robert P. Sharp to Penrose Medal Committee, Geological Society of America, 3 December 1991, box 9, folder 15, Papers of Robert P. Sharp, California Institute of Technology.
8
J. Hoover. Mackin, “Rational and Empirical Methods of Investigation in Geology,” in The Fabric of Geology, ed. Claude C. Albritton, Jr. (Stanford, CA: Freeman, Cooper and Co., 1963), 135–163.
9
→Luna B. Leopold, Water and the Conservation Movement (Washington, DC: US Geological Survey Circular, 1958).
→Luna B. Leopold, A View of the River (Cambridge, MA: Harvard University Press, 1994).
10
Durham, NC: Duke University Press, 2015More Info →
11
Kristoffer Whitney and Melanie A. Kiechle, “Introduction: Counting on Nature,” Science as Culture 26, no. 1 (2017): 1–10.
12
→Aldo Leopold, A Sand County Almanac, and Sketches Here and There (New York: Oxford University Press, 1949).
→Julianne Lutz Newton, Aldo Leopold’s Odyssey (Washington: Island Press/Shearwater Books, 2006).
13
→Hans Reichenbach, Experience and Prediction: An Analysis of the Foundations and the Structure of Knowledge (Chicago: University of Chicago Press, 1938).
→Naomi Oreskes, “A Context of Motivation: US Navy Oceanographic Research and the Discovery of Sea-Floor Hydrothermal Vents,” Social Studies of Science 33, no. 5 (2003): 697–742.