2. Background

Aldo Leopold (1887–1948) was a 20th-century hunter, forester, wildlife manager, ecologist, conservationist, and professor. Among philosophers, he is best known for his posthumously published book A Sand County Almanac (ASCA)—in particular, the chapter entitled “The Land Ethic.” Yet during his lifetime he produced more than 500 published and unpublished works, many written for a scientific audience rather than a general audience; ASCA was written for a general (layperson) audience. Understanding Leopold requires going beyond ASCA.

Leopold’s approach to science was value-laden in a number of respects, so that by the end of his life he had identified various entities and properties to be of value: land communities, members of land communities, land health, and the factors necessary to sustain land health (species interactions/interdependence, biodiversity, soil health). Moreover, he thought that science and ethics could not be separated (Meine Reference Meine1987, 177):

He also advocated incorporating values into policy decisions. For example, he suggested an approach that would “reorganize and gear up the farming, forestry, game cropping, erosion control, scenery, or whatever values may be involved so that they collectively comprise a harmonious balanced system of land-use” (Leopold Reference Leopold, Baird Callicott and Eric1935/1999, 47; emphasis added).

In one of his most famous essays, “Thinking Like a Mountain,” also in ASCA, Leopold encapsulated in several short paragraphs something that in fact took him a lifetime to realize (Flader Reference Flader1994, Meine Reference Meine2010):

How does someone go from wanting to kill every wolf to mourning the death of one wolf? I use Leopold’s trajectory of this change in his thinking to illustrate how his evaluative beliefs coevolved with his descriptive beliefs over time. But first, a couple of caveats. This will be a high-level overview, lacking much of the detail. Moreover, note that I primarily follow only one strand of Leopold’s thinking—other related ones would be, e.g., his changing views of land health, the nature of a biotic/land community, and land management.

3. Trajectory of Leopold’s views on wolves

The following chronology was developed using material from Flader (Reference Flader1994) and Meine (Reference Meine2010) in conjunction with primary sources from Leopold.

1887–1904 (youth): From his father Carl Leopold, Aldo Leopold learns to appreciate and enjoy the natural world; his father also impresses upon him a hunter’s ethics.

1909: Leopold begins work as a forester in the Apache National Forest.

1918: Broadening his research interests from forestry to “game management,” Leopold writes of the value of biodiversity in trees and in “game” and starts to become concerned about land sickness, but predators/wolves play no role in his views of either (Flader Reference Flader1994).

1920: He advocates extirpating wolves in New Mexico in order to provide more deer for hunting: “It is going to take patience and money to catch the last wolf or lion in New Mexico. But the last one must be caught before the job can be called fully successful” (Leopold Reference Leopold1920). Leopold thinks that the absence of predators means more deer to hunt. In this same year, Leopold first uses the word “ecology” in print, a fairly new term that he had learned about from his professional reading (Meine Reference Meine2010), but he is seemingly not (yet) seeing the extirpation of wolves in an ecological context.

1925: Leopold states, “To facilitate ecological studies, the introduction of exotics should be carefully avoided … For the same reason, it is important to avoid the extermination of predators, but there is no danger of this as yet” (quoted in Meine Reference Meine2010, 242; emphasis added). So, he has eased off from wanting to eliminate wolves, but not (yet) from predator control.

1927: Leopold receives a report of overabundance of deer and damage to forage in the Gila Wilderness in New Mexico. There were also reports of dramatic increases in the number of deer and effects on the supply of forage in the Kaibab National Forest in Arizona. He seems to have assumed these were special cases and not “cause for alarm” (Flader Reference Flader1994, 120). As a trained, professional forester, Leopold would have been attentive to data about damaged forests.

1929: He receives more data from the Gila, now showing that the buck:doe ratio was 1:4; he speculates that many of the does were barren and that killing off too many of the large predators might be one factor in that, given that small predators like coyotes primarily kill the young deer. In response, he recommends focusing on killing coyotes and letting the lions alone for a while.

1929–1930: He receives more bad information about deer from the Gila, including reports of unhealthy deer. There is disagreement over what to do, with some arguing for reducing the size of the deer herd, but Leopold refrains from taking sides. Some argue that it was the cattle, not the deer, that were the problem.

1935: This seems to have been a turning point, i.e., the point where it becomes fully clear to Leopold that the absence of big predators like wolves is a major problem. During a trip to Germany, Leopold sees severely damaged forests and deer with nutritional deficiencies, yet the forests had been conscientiously managed. That the forests were managed eliminates other possible explanations for the deterioration, such as the explanations that he invoked in earlier cases of problems with deer herds. The size of the deer herd was the only remaining plausible explanation. Around the same time, foresters in northwestern Wisconsin found large deer population sizes with overbrowsing and starvation of deer; they thus recommended herd reduction. Concerning Germany, he writes:

By 1936, when he visits Chihuahua, Mexico, Leopold sees wolves as essential for land health in that region. As the years pass, he learns about more and more problems in predator-less land communities.

1944: Leopold writes a book review that criticizes the book’s lack of mention of

By 1948 (the end of his life), wolves are incorporated into Leopold’s land ethic. He sees them as having rights in virtue of being members of the land community as well as essential for the health of the whole. He writes:

In sum: Leopold goes from thinking that wolves should be extirpated; to thinking that predator control was OK as long as predators weren’t extirpated, but mainly because predators were needed for ecological study; to explaining away data that were showing a connection between large deer populations and damaged forests, at best acknowledging predators as one factor among many; to finally being confronted with data so incontrovertible that he believes a lack of predators to be the cause of deer population explosions that lead to damaged forests and starving deer; to thinking that wolves were essential for land health; to admitting in print that he had been tragically mistaken in thinking that there couldn’t be too much “game” and in calling for the extirpation of predators; and finally, by the end of his life, seeing wolves and other predators as members of the community who should continue as a matter of biotic right. This is a tectonic shift in values, but it is a shift that occurs gradually throughout the course of his lifetime, with data playing a central role in that shift.

It all began with valuing nature experiences and hunting. These are the values that first led him to think that attempting to extirpate wolves was a good idea, a belief he did not hold about non-predators. The belief about extirpating wolves is a descriptive belief about what actions would yield more deer and a better hunting experience, but it is a descriptive belief tinged with evaluative beliefs about the value of hunting and what makes for good hunting. His values perhaps caused him to hold on to that descriptive belief longer than he ought to have, even when the evidence was telling him otherwise. But finally the data became incontrovertible. And when that happened, he changed his descriptive beliefs about killing wolves, and, in short order, also changed his evaluative beliefs about wolves in a way that was more in line with his values concerning other animals and that also preserved his hunter’s values. Thus, both the data and his values played a role in his new descriptive beliefs. Moreover, his beliefs about how to preserve (descriptive), and the importance of preserving (evaluative), species interdependencies and land health were expanded.

4. Implications of Leopold’s trajectory

This trajectory shows how Leopold’s evaluative beliefs and descriptive beliefs changed over time. As a deer hunter, Leopold valued certain things (being out in nature, large deer populations) and not other things (wolves), but gradually and ultimately came to believe valuing deer hunting was undermined by disvaluing wolves. In light of his experiences, he came to see wolves, deer, and all the members of the land community as valuable, and essential parts of a valuable whole, a belief that is both descriptive and evaluative. As historian Susan Flader argues: “Leopold’s thinking was shaped by the land itself, and by his changing perception of it … When one looks for critical junctures in his thinking, one finds them as often as not associated with some new field experience.” (Flader Reference Flader1994, 35; emphasis added) ^{Footnote 3}

Elizabeth Anderson (Reference Anderson2004) claims that there can be a bidirectional influence between “fact judgements” and “value judgements.” ^{Footnote 4} That is, she maintains, facts can count as evidence for value judgments, and value judgments can help us see certain facts. To which I add: this bidirectional influence can be ongoing—there can be coevolution of judgments—over the course of years. Leopold’s valuing of the outdoors led him into forestry, but forestry didn’t fully capture all he found valuable in the outdoors, eventually leading him to game management. But game management turned out to be a difficult issue to address—in fact, not just for deer/wolves, but more generally (see Meine Reference Meine2010). Eliminating all the predators did not have the expected results. The data could be explained away for many years, but eventually Leopold was forced to accept that more deer did not lead to better hunting—it led to denuded forests and starving deer. During the same time, Leopold begins to emphasize the value of a healthy land community, and eventually fits wolves into that picture. And that eventually becomes the value of wolves in and of themselves, regardless of their benefits to others.

In short, Leopold’s trajectory exhibits a bidirectional influence between descriptive beliefs and evaluative beliefs over decades—a coevolution. Leopold stated that the land ethic is a “product of social evolution” (1949, 225). Perhaps that reflects the evolution of his own ideas, with descriptive beliefs modifying evaluative beliefs (and vice versa), analogous to the way that bees modify flowers (and vice versa) over the course of many years. Of course, social (i.e., cultural) evolution is famously much faster than biological evolution and the mechanisms are not identical; I use “coevolution” in a metaphorical sense here.

Seeing a case where evaluative beliefs and descriptive beliefs coevolve over time allows us to see that refinements of beliefs need not be large or sudden. This makes the refinements more psychologically plausible even if human agents aren’t reliably open-minded (cf. Yap Reference Yap2016). That is, there is no presumption here that new data will cause a person to experience a large change in their descriptive or evaluative beliefs, or any change in beliefs at all. The presumption is rather the much more modest one that they can cause such changes, gradually and over time.

Anderson (Reference Anderson2004) further suggests that some values may be systematically more epistemically fruitful than others. Again, Leopold’s trajectory illustrates this well. Leopold’s early values turned out to be epistemically unfruitful. His valuing the experience of deer hunting a large population of deer turned out to undermine not only deer hunting but also the health of the deer and the health of the land communities that they were members of. His early values led him to advocate poor management policies; he spent his later years fighting to correct those policies as his evaluative beliefs and descriptive beliefs changed.

However, unlike Anderson, I resist the idea that “[t]he primary evidence for being mistaken about a standard of appraisal [for a value judgment] is itself tied to emotional experience” (Anderson Reference Anderson2006, 5; emphasis added). I think that emotional experiences can serve as evidence for value judgments, so I am not fully objecting to Anderson’s view here, but Leopold’s life shows that other experiences, like field experiences, can do the same. This is not to say that Leopold’s emotional experiences were irrelevant to his values, just that they did not seem to be the driving forces behind his changes in values. Rather, the available evidence suggests that the driving forces were typically his field experiences and data exposure more generally. ^{Footnote 5}

Sharyn Clough’s account of the relation between evidence and values—that no intermediary, emotional or otherwise, is needed between evidence and values—better fits the lessons of Leopold’s trajectory. Clough asserts that:

Leopold’s early valuing of large deer populations and disvaluing of wolves indeed exhibited values with empirical content. His values presumed that the best way to preserve the deer-hunting experience was by killing predators, which was entangled with a disvaluing of predators; early on, Leopold referred to wolves as “vermin,” saying that “advisability of controlling vermin is plain common sense, which nobody will seriously question” (Leopold Reference Leopold and Meine1919/2013, 201; emphasis added). But those values—that “common sense”—turned out to have poor empirical support. Later evidence overturned them for more empirically supported values: that the deer-hunting experience cannot be maintained without large predators like wolves, and, eventually, for the view that wolves should be allowed to continue as a matter of biotic right and that they are essential for land health in many cases. ^{Footnote 6}

Clough further maintains “that scientists, as with all of us, can increase the objectivity of research by adopting a pragmatically-inclined appreciation of the fallible, inductive process by which we gather evidence in support of any of our beliefs, whether they are described as evaluative or descriptive” (2020, 7; emphasis added). Again, this fits Leopold’s story. The process was certainly a fallible one; early evidence concerning the failures of “predator control” didn’t change Leopold’s mind, and then when it did start to change his mind, it was only incrementally. But eventually, inductively, Leopold came to see things differently. Pragmatically, Leopold’s early values simply didn’t work. His later values, which shifted toward the value of biodiversity and soil health for the value of land health more generally, worked better—even as these are still fallible, as all evaluative and descriptive beliefs are.

Given that the process by which we gather evidence in support of our beliefs is a fallible one, it should not be surprising that there are those who still do not accept Leopold’s view that wolves are and should be treated as members of the community (a belief that has both evaluative and descriptive aspects). Indeed, as a recent article notes, “a common belief about large predators is that they compete with hunters by decreasing ungulate populations through additive mortality” (Trump et al. Reference Trump, Knopff, Morehouse and Boyce2022, 2), a belief that the authors seek to debunk through their study of elk in Alberta over a 26-year period. This belief is sometimes conjoined with a belief that predator “control” ought to be instituted, even though most studies show it not to be efficacious or cost-effective (Miller et al. Reference Miller, McLellan and Derocher2013, 390). Evaluative beliefs can become entrenched and be hard to change—but so can descriptive beliefs. The point here is that they often go together. Perhaps the best that can be done is to try to show people that given certain beliefs they have (e.g., about the value of hunting ungulates), other beliefs that they have (e.g., about the disvalue of wolves) are simply not consistent—that their complete set of values doesn’t work, necessitating change. There is no guarantee that that process will succeed, but Leopold’s life trajectory shows that it is possible that it will.

Leopold’s evaluative beliefs, his values, were tested—well-tested—as were his descriptive beliefs. That is, they evolved over time as they were tested against the evidence of his experiences, in ways that go beyond what I have space to discuss here. To the extent that these tested beliefs still underlie much of conservation biology—and I think they do to a significant extent—they put conservation biology on a firm empirical footing. New empirical findings may change those values to some extent, and modified values may change descriptive beliefs in turn, but a proper understanding of the relation between evaluative beliefs and descriptive beliefs shows that the value-ladenness of today’s conservation biology can be an asset rather than a liability. Some of these values are well supported by evidence and relevant to projects of importance. They thus help strengthen the empirical adequacy of conservation biology.

5. Summary and conclusions

Leopold’s evaluative and descriptive beliefs about wolves and deer (and biodiversity and land health) not only influenced each other, they did so repeatedly over the course of his career, each modifying the other in light of his experiences in the world, in what I am calling a coevolutionary process. My analysis of Leopold supports Anderson’s (Reference Anderson2004) account of the bidirectional influence of “fact judgements” and “value judgements.” It also supports Clough’s (Reference Clough2020) account of the direct effect of experiences in the world on both evaluative and descriptive beliefs in a fallible, inductive process. And it extends both accounts by showing that these influences can occur repeatedly over an extended period, so that the two types of beliefs coevolve, being tested and refined over time. Although it is a single case, it shows that such coevolutions of belief are possible and worth looking for elsewhere.

To the extent that the values articulated by Leopold and Soulé in turn still underlie conservation biology, they provide a firm grounding for future findings in the field. Soulé was right that such value judgments can’t be “proven,” but they can be shown to be fruitful, sustainable guides to policy and action, even as they continue to be modified over time.