It takes . . . a mind debauched by learning to carry the process of making the natural seem strange so far as to ask for the why of any instinctive human act.
—William James (1887)
Adam Smith’s account of the moral sentiments resonates with modern themes in evolutionary biology. His distinction between our reasons and the reasons for these reasons recalls evolutionary biologists’ distinction between explanations of mechanisms and explanations of why those mechanisms have evolved. This distinction is commonly accredited to Ernst Mayr (1961) and his separation of proximate and ultimate causes, although I think this interpretation is a substantial misreading of Mayr’s text. This chapter originated as my response to an invitation to comment on the distinction between proximate and ultimate causes. Whether this distinction clarified or obfuscated understanding of the evolutionary process had become a point of passionate contention in the philosophy of biology. Some material on pre-nineteenth-century uses of “proximate cause” and “ultimate cause” were not considered germane by the soliciting editor and were deleted at his request but have been restored in this version. My purpose of including this chapter in the current volume is to illustrate that concepts of cause are not as straightforward as most biologists think, to illustrate the aversion many biologists have for teleological language, and to illustrate the confusions that can result from the ambiguities of words.
Ernst Mayr (1961) was a staunch advocate of historical explanations in biology: ultimate causes were “causes that have a history.” He believed one could not fully understand biological processes without understanding their evolutionary history. A parallel argument can be made for the meanings of words. Meanings evolve by mutation, semantic drift, and competition among alternatives. At any one time, variant definitions may be adopted by different individuals, or by the same individual in different contexts, and the frequency of variants may change over time. Which senses prevail may shape the direction and outcome of scientific and philosophical debates. Protagonists want others to see the world in their terms. One benefit of addressing how meanings have evolved is to step back from arguments about the “true,” “correct,” or “real” meanings of words.
The distinction between proximate and ultimate causes in evolutionary biology has variously been interpreted as a distinction between causes in the present and causes in the evolutionary past or between explanations of mechanism and explanations of adaptive function. Everyone seems to agree what are proximate causes. These are Aristotelian efficient causes. Differences of opinion center around the understanding of ultimate causes. Ambiguity in the meaning of “ultimate cause” has existed for centuries. “Ultimate cause” could refer either to the first in a series of efficient causes or to a final cause, often conceived as coming at the end of a series of events. The current ambiguity—whether ultimate causes are historical explanations or functional explanations—can be considered a descendant of this older ambiguity in the meaning of ultimate cause. Current debates would be clarified by distinguishing between these different senses of ultimate cause.
The Oxford English Dictionary is a good place to begin. The adjectives proximate and ultimate are derived from Latin verbs for “to draw near” and “to be at the end.” The primary definition of proximate is “Coming immediately before or after in a chain of causation. . . . Freq. in proximate cause. Opposed to remote or ultimate.” The primary definition of ultimate is “Of ends, designs, etc.: Lying beyond all others; forming the final aim or object.” The earliest use of both adjectives is dated to the mid-seventeenth century. Of particular significance, “ultimate cause” has long been associated with “final aims,” and “proximate cause” has been variously opposed to “remote cause” or “ultimate cause.”
Proxima causa and ultima causa were used in scholastic Latin for centuries before their adoption into English as proximate and ultimate cause. In De principiis naturae, Thomas of Aquinas presented an Aristotelian distinction between prior and posterior causes:
We must understand that proximate cause means the same as posterior cause, and remote cause the same as prior cause. Accordingly, these two divisions of causes—prior and posterior, and remote and proximate—signify the same thing. Furthermore, we should observe that the more universal cause is always called the remote cause, while the more particular cause is called the proximate. For example, the proximate form of man is his definition, that is, rational mortal animal; but animal is more remote, and substance even further removed. For all superior things are forms of inferior ones. Similarly, the proximate matter of a statue is bronze, while the remote matter is metal, and the more remote is body. (1965, 23)
Remote causes explain proximate causes, but not the reverse. In the above passage, Thomas illustrates the distinction between proximate and remote causes by using hierarchies of formal and material causes in which priority is generality: general causes are remote and specific causes proximate. However, the distinction also applies to chains of efficient causation where priority is temporal—earlier events cause later events—and to final causes where proximate ends are means to higher goals.
In any causal chain or hierarchy, the most remote cause, the cause without a prior cause, was the ultimate cause. In Summa contra gentiles, Thomas used Aristotelian arguments against infinite causal chains to demonstrate the necessity of ultimate causes. He reasoned that there must be a first unmoved mover in chains of efficient causation and a first cause that is an end in itself in chains of final causes (Aquinas 1975, chap. 13, 37–38). For Thomas, the unmoved mover and ultimate end (ultima finis) were one and the same. In the beginning is the end.
Spinoza similarly distinguished between proximate and remote causes. In his Short Treatise on God he averred
God is the proximate cause of the things that are infinite and immutable and which we assert to have been created immediately by him, but, in one sense he is the remote cause of all individual things. (2002, 51)
Proposition 28 of his Ethics argued for infinite chains of efficient causes in the realm of individual (finite) things. The Scholium to that proposition states:
It follows, firstly, that God is absolutely the proximate cause of things directly produced by him. . . . It follows, secondly, that God cannot properly be said to be the remote cause of individual things. . . . For by “remote cause” we understand a cause which is in no way conjoined with its effect. But all things that are, are in God, and depend on God in such a way that they can neither be nor be conceived without him. (2002, 51)
Spinoza seems to change his mind about whether God is a remote cause of individual things. In Proposition 28, God is eternally present, proximate and not remote. Spinoza categorically rejected final causes, “all final causes are figments of the human imagination” (2002, 59). Spinoza denied final causes even to God, because everything is in God and if God acted with an end in view he would necessarily be seeking something he lacked.
Proximate cause also has a rich legal history, especially in the law of torts. Francis Bacon’s (1596) first maxim of the law was In jure non remota causa sed proxima spectatur. If every cause has prior causes, then the law pragmatically concerns itself only with the immediate or proximate cause rather than engage in an investigation of remote causes. In summary, causes were ordered by metaphoric or actual time (prior and posterior) or by metaphoric distance (proximate and remote). These axes were orthogonal to distinctions among efficient, material, formal and final causes. Ultimate causes could be invoked for all four Aristotelian categories of cause.
It is not my intention to review the diverse meanings and nuances of ultimate and proximate causes in the nineteenth century. I will limit myself to some illustrative examples from medicine and the philosophy of Herbert Spencer. In medicine, proximate (or immediate) causes of disease were distinguished from remote causes with occasional references to ultimate (or primary) causes. The second volume of Erasmus Darwin’s Zoonomia presented a classification of diseases according to proximate causes:
Thus in the cramp of the calves of the legs in diarrhœa, the increased sensorial power of association is the proximate cause; the preceding increased action of the bowels is the remote cause; and the proximate effect is the violent contractions of the musculi gastrocnemii; but the pain of these muscles is only an attendant symptom, or a remote effect. (1818, 361)
John Chapman similarly addressed the causes of diarrhea. It could arise
from the irritation of dentition, from irritating ingesta, from drinking impure water, from inhaling noxious gases, from ulceration of the bowels (as in phthisis), and the symptomatic diarrhœa of numerous diseases” but “however diverse the primary cause of the disease, the proximate cause is always the same . . . viz., hyperæmia of the spinal and sympathetic nervous centres. (1865, 14)
Here, the proximate cause was the final mechanism, elicited by diverse primary causes, that resulted in shared symptoms.
In the context of disease, ultimate causes were physical causes that occurred before proximate causes. Thus, in a discussion of potato blight, “the Peronospora is undoubtedly the proximate cause of the disease, for the ultimate cause we may look to a very different set of circumstances. . . . The attacks of the parasite may be favoured by special climatal or other conditions” (“The Potato Disease,” 1872). Similarly, “a fall early in January last resulting in a fractured thigh was the ultimate cause of his death, which occurred on September 5” (“Rudolf Ludwig Karl Virchow: Obituary,” 1902). By contrast, the “ultimate cause of death” in contemporary medicine is often the cause at the end of the series, immediately before death, rather than the remote precipitating cause at the beginning of the series. For example, “The ultimate cause of death in patients with sepsis is multiple organ failure. Typically, patients will first develop a single organ failure . . . and then if the disease remains unchecked, will progressively develop failure of other organ systems” (J. Cohen 2002).
Herbert Spencer described “the necessary antagonism of individuation and reproduction” as ensuring “the final attainment of the highest form of [racial] maintenance” and as leading to “the ultimate disappearance of the original excess of fertility” (1852, 501). He continued: “From the beginning, pressure of population has been the proximate cause of progress. . . . It compelled men to abandon predatory habits. . . . It forced men into the social state. . . . And after having caused, as it ultimately must, the due peopling of the globe, and the bringing of all of its habitable parts into the highest state of culture . . . after having done all this, we see that the pressure of population, as it gradually finishes its work, must gradually bring itself to an end.” Proximate cause, in this passage, is the means that brings about a predestined or higher end. In a similar vein, the locomotive engine was “the proximate cause of our railway system [that] has changed the face of the country, the course of trade, and the habits of the people” (Spencer 1857, 481).
For George Stokes (1887), “the highest aim of physical science is, as far as may be possible, to refer phenomena to their proximate causes.” However, beyond a certain point “Science conducts us to a void she cannot fill.” Spencer had earlier addressed this void. In First Principles, he equated Ultimate Cause (capitalized) with the Unknowable “that through which all things exist” (Spencer 1867, 108–114). However, when he turned to the knowable, and the instability of homogenous bodies, certain slight differences among units were doubtless the “proximate cause” of heterogeneity but the “ultimate cause” (uncapitalized) was unequal exposure of the parts to incident forces (1867, 424).
In summary, proximate causes were physical causes. “Ultimate cause” could refer to a physical cause that occurred before a proximate cause or to a final cause. Finally, in preparation for my discussion of Mayr’s distinction between proximate and ultimate causes, it is worth noting that efficient causes were often associated with how questions and final causes with why questions. I will give two examples from the nineteenth century of this contrast. First, Johann Peter Eckermann reports Goethe stating: “Die Frage nach dem Zweck, die Frage Warum? ist durchaus nicht wissenschaftlich. Etwas weiter aber kommt man mit der Frage Wie?” (1836, 283). In my imperfect translation: “The question of purpose, the question why? is not at all scientific. But one can get a little further with the question how?” Second, Charles Kingsley wrote in First Course of Earth Lore for Children: “But of one thing I must warn you, that you must not confound Madam How and Lady Why. Many people do it and fall into great mistakes thereby” (1873, 4). He identified Madam How with Nature and her actions with Fact (348). Her beautiful mistress, Lady Why remained unidentified “though she has a Master over her again—whose name I leave for you to guess” (3).
The attempts to answer the questions “Why” and “How”—“To what end?” and “In what way?”—by no means interfere with each other. . . . There are always these two questions to be answered with reference to any one natural phenomenon, and both must be answered if the facts are to be fully understood.
—Edward Poulton (1908)
Ernst Mayr presented his “Cause and Effect in Biology” to the Hayden Colloquium on Scientific Method and Concept at MIT during the academic year 1960–1961 (Beatty 1994; Lerner 1965). A version was published in Science (Mayr 1961) and subsequently in the proceedings of the Hayden Colloquium (Mayr 1965). The two versions are highly similar but not identical. The 1961 version (which I will use in my discussion) is commonly presented to students as distinguishing how questions (proximate causes) from why questions (ultimate causes) with why interpreted as the question of adaptive purpose, but this is a misinterpretation of Mayr’s text.
Mayr (1961) was not the first biologist to attach how to proximate causes and why to ultimate causes. A decade earlier, John H. Mullahy had written:
Scientists as such are not prepared to answer the ultimate questions of why there is a universe and why the universe evolves. The scientist is only asked to describe the process; his profession demands only that he tell us how the universe evolves. When scientists pause in their inquiry into the proximate explanation of things and brashly hybridize their professional status by taking the philosophical bit between their teeth, the sterility of their effort serves only to remind us of other hybrids more accustomed to the bit. (1951, 20)
Similarly, Claude Wardlaw wrote that biology could aspire to an understanding of “the ultimate and proximate causes of evolution and of the mechanism of life and morphogenesis,” but the “reason for evolution, i.e. the why of evolution is generally regarded as belonging to a branch of learning other than biology” (1952, 463). Here, Wardlaw used “ultimate cause” to refer to remote physical causes. Like Mullahy (and before him Goethe and Kinglsey), Wardlaw considered reasons why to lie outside the domain of science. It should be noted that Wardlaw’s and Mullahy’s why is the what for of the evolutionary process itself, not the what for of the adaptive products of evolution.
One must consider Mayr’s “political” motivations to understand the goals of “Cause and Effect in Biology.” An important motivation was to defend the place of systematics and evolutionary biology in the academy against triumphalist molecular biology (Beatty 1994; Dietrich 1998). The criticisms against which Mayr wished to defend his discipline are not difficult to divine, both from his text and because evolutionary biologists still encounter the same criticisms. Evolutionary biology, it is said, smuggles in unscientific, teleological concepts and is less rigorous and predictive than the “hard” sciences.
At the same time, Mayr wanted to disown any connection between evolutionary biology, correctly delimited, and various vitalistic theories that were current in his youth. His opening paragraph addresses the theories of Driesch, Bergson, and Lecompte du Noüy, and its last paragraph concludes: “The complexities of biological causality do not justify embracing nonscientific ideologies, such as vitalism or finalism” (1961, 1506). Given this prominence, one might be tempted to conclude that vitalism, rather than proximate and ultimate causes, was Mayr’s principal target and that he wanted to drive a stake through the heart of theories that, at the time of his writing, were already feeble if not dead. His reason for flaying the dead horse of vitalism was tactical. Vitalistic theories had been an inevitable reaction to “Descartes’s grossly mechanistic interpretation of life” (1501). Mayr wanted to argue that functional biologists’ concepts of causation, derived from the physical sciences, were impoverished and inadequate for understanding the living world, whereas evolutionary biologists had a richer view; but he needed to defend himself against the accusation that he was invoking some entelechy or elán vital in addition to physical causation.
“Cause and Effect in Biology” is structured around three aspects of causality: explanation of past events, prediction of future events, and teleology. Mayr first distinguished two broad disciplines, functional and evolutionary biology, with different explanatory goals and different causal concepts. He identified the functional biologist’s question as how and the evolutionary biologist’s question as why. The functional biologist was concerned with immediate causes, whereas the evolutionary biologist was concerned with “causes that have a history.” Thus, “proximate causes govern the responses of the individual (and his organs) to immediate factors of the environment, while ultimate causes are responsible for the evolution of the particular DNA code of information with which every individual of every species is endowed” (1503).
Because Mayr’s “ultimate cause” is commonly interpreted as the explanation of selective value or purpose, it is worth quoting his explicit disavowal of this interpretation:
When we say “why” we must always be aware of the ambiguity of this term. It may mean “how come?,” but it may also mean the finalistic “what for?” It is obvious that the evolutionary biologist has in mind the historical “how come?” when he asks “why?” (1502)
What did Mayr mean by “proximate” and “ultimate” causes and why did he choose these terms? More than once, Mayr states that proximate causes are immediate whereas ultimate causes are historical. Therefore, ultimate causes are presented as temporally prior to proximate causes. Mayr clearly did not believe that evolutionary causes were without prior cause, so why choose “ultimate” rather than, say, “remote”? This was politics. Proximate causes are more salient than remote causes, but ultimate valorizes the evolutionary over the merely proximate.
Mayr’s (1961) treatment of teleology is revealing. He explictly disavowed “What for?” because of its teleological overtones. The idea that evolution had a goal or purpose was the feature of vitalistic theories to which Mayr most strongly objected. Goal-directed behaviors are present in biology, but these are caused by the playing out of evolved genetic programs. By implication, how these programs are expressed is the province of functional biologists, but the origin of these programs is the realm of evolutionary biologists. “The development or behavior of an individual is purposive, natural selection is definitely not” (1504). Mayr subtly shifts the stigma of teleological thinking from evolutionary to functional biology. Although evolution was not goal-directed, Mayr recognized that natural selection produced organisms with apparent purposes. Rather than describe goal-directed behaviors as teleological, he preferred to use the adjective teleonomic and to restrict this term “rigidly to systems operating on the basis of a program, a code of information” (1504).
Mayr (1969) returned to these themes in “Footnotes on the Philosophy of Biology.” “Vitalism,” he declared, “is now dead, as far as biologists are concerned. . . . The enormous complexity of biological systems, the historical nature of all organisms, and the fact that organisms contain a historically evolved genetic program make them so totally different from nonliving objects that generalizations derived from nonliving objects are in most cases meaningless or trivial when applied to phenomena of life. . . . Purposive processes and behaviors that owe the adaptive nature of their response to genetic programming, may be designated as teleonomic. That part of Aristotle’s teleology is not acceptable according to which the universe as a whole is programmed to evolve harmoniously, ultimately resulting in a harmonious kosmos. This is teleology sensu stricto and for this no substantiation has been found in any area of science” (197–202).
Mayr’s understanding of these issues evolved. Mayr (1974) discussed the relation between teleology and teleonomy: “A teleonomic process or behavior is one which owes its goal-directedness to the operation of a program” (98). The significant change in the reworded definition was that he deleted “system,” which he now saw as static and not dynamic. He was now unwilling to concede an eye, or a torpedo before it is fired, as being teleonomic. “Goal-directed” was not the same as “purposive.” “Teleonomy” was not synonymous with “adaptation.” He had also shifted ground on the evolutionary why. What and how were now considered adequate for the physical sciences, but no biological explanation was complete without an answer to why. “It is necessary for the completion of causal analysis to ask for any feature, why it exists, that is what its function and role in the life of the particular organism is” (108). Asking why “demands asking for the selective significance of every aspect of the phenotype” (109). This is what for? rather than how come? Mayr (1992, 131) wrote: “Adaptedness is an a posteriori result rather than an a priori goal-seeking. For this reason the word teleological is misleading when applied to adapted features.” Mayr (1993, 94) recognized the historical association of ultimate causes with natural theology and wrote: “In order to shed the historical impediment of the term ultimate, I have used in most of my recent papers the term ‘evolutionary’ instead of ultimate causation.”
The older Mayr was notable for dogmatic, one might say ex cathedra, pronouncements. A personal anecdote may be of interest to historians of evolutionary biology. Early in 1999, Mayr summoned me to his office. “We are going to write a paper together,” he declared. “My English friend,” he hesitated—“John Maynard Smith,” I suggested—“Yes,” he confirmed, “is wrong: animals do not play games.” This was just after the announcement that Mayr was to share the Crafoord Prize with John Maynard Smith and George Williams, so I asked him how he felt about sharing the prize with Maynard Smith. The prize was “well deserved,” but not for Williams: “Everything he has ever done has been wrong!”
Mayr’s distinction between proximate and ultimate causes was widely adopted by evolutionary biologists, but largely ignored by functional biologists. Perhaps some of the attraction for evolutionary biologists was the connotations of “ultimate” and “proximate.” It is easy to imply that ultimate causes are more important than proximate causes. But this invidious comparison certainly did not endear “ultimate” to functional biologists who were interested in proximate causes. “Ultimate cause” has been labeled “simply pretentious” (Francis 1990) and “disciplinary chauvinism” (Dewsbury 1999). Thirty years after Mayr (1961) first contrasted biological explanations in terms of proximate and ultimate causes, Beatty wrote that “more than minor disagreement with his insistence on the proximate/ultimate distinction would be heretical” (1994, 352). However, recent exchanges in the philosophy of biology suggest a doctrinal dispute has broken out, in which papal authority is challenged by latter-day Luthers (Laland et al. 2013b; Thierry 2005) and defended by contemporary Johns of the Cross (Dickins and Barton 2013; Gardner 2013).
These recent debates about whether the proximate–ultimate distinction is useful, outdated, or pernicious have involved much arguing at cross-purposes because proximate and ultimate causes have come to conflate two distinctions. The first is a distinction between immediate and historical causes. The second is a distinction between mechanism and adaptive function. Mayr (1961) emphasized the first distinction, but most evolutionary biologists who adopted his terminology emphasized the second. This has created a semantic morass in which some critics of Mayr’s distinction employ a definition closer to Mayr’s intended meaning than do most supporters of the distinction. Two criteria are often invoked to determine the “correct” definition of a term. One is historical: what did Mayr mean by “ultimate cause”? The other is majoritarian: what do most individuals in community x mean by “ultimate cause”? Unfortunately, these criteria suggest different answers to the “correct” meaning of “ultimate cause.”
Ambiguity in “ultimate cause” arose from an ambiguity in why? that Tinbergen (1968) avoided by separating questions of survival value from evolutionary history. Mayr explicitly disavowed the finalistic what for? in favor of the historical how come? but many evolutionary biologists wished to defend what for? against the mechanistic how? What for? is exclusively concerned with natural selection, but how come? encompasses additional historical factors. Close attention to whether authors identify ultimate causes with what for? or how come? clarifies their positions and illuminates causes of misunderstanding in current debates. Some see the relevant distinction as predominantly temporal (of immediate versus historical causes):
This conceptual dichotomy is a deeply engrained habit of thinking and is characterized by the belief that aspects of development are determined by either (a) events which occurred earlier in the development of the individual, or (b) preontogenetic factors which operated on the ancestors of the individual. (Lickliter and Berry 1990, 349)
Proximate explanations focus on causes in the present; evolutionary explanations focus on how the present has been shaped by events in the past. (Hochman 2013, 593)
Proximate causes are immediate, mechanical influences on a trait. . . . Ultimate causes are historical explanations. (Laland et al. 2013a, 720)
From this perspective, the distinction between proximate and evolutionary causes is seen as a false dichotomy, without a principled distinction between how and how come. By contrast, defenders of a proximate–ultimate distinction usually focus on the difference between how and what for and implicitly or explicitly place how come on the proximate side of the ledger:
To understand behavior . . . it is vital to distinguish between the “proximate” cause (the physiological mechanism) and the “ultimate” cause (the evolutionary “goal”) of the behavior in question. (Burnham and Johnson 2005, 124)
The explanatory utility of an ultimate account is not solely in a detailed phylogeny, capturing a particular trait’s ancestry and shifting phenotype over time; it is the exposure of function, of why a given trait is as it is. . . . Historical accounts are not in any sense default ultimate accounts, they . . . can be understood in purely proximate terms. (Dickins and Barton 2013, 749)
Two issues have become enmeshed. The first concerns what efficient causes are important in evolutionary change. Do developmental mechanisms play a role? Does organismal evolution alter the environment and subsequent selection? Few would dissent from an affirmative answer to either question. The problem is that some define these as proximate and others as ultimate questions.
The second issue concerns the role of teleological reasoning and language in evolutionary biology. From this perspective, how and why separate questions of mechanism from questions of function. This separation is the evolutionary descendant of the venerable distinction between efficient and final causes, with final causes interpreted in a strictly Darwinian sense. Some are comfortable with Darwinized final causes, but others believe they have no place in science.
Cause-talk and function-talk are not simply different vocabularies, they are incommensurable. There is no common currency in which causal and functional explanations can be cashed in. There is no common currency for proximate and ultimate causes. There are no ultimate causes. (Francis 1990, 413)
The mechanisms behind evolutionary processes leading to novelty can be explained in the absence of ultimate (final) causes, but not in the absence of proximate (efficient) causes. (Guerrero-Bosagna 2012, 285)
“Why is A?” has two different, if related, meanings. If we are dealing with a conscious (or programmed) agent, we may ask why the agent took action A, seeking explanation of the agent’s reason for action A. Absent such an agent, “why is A?” morphs into a “how?”-question from “how does A come about?” . . . Thus “why?” questions concerning natural selection actually are “how?” questions. (Watt 2013, 760)
Debate would be clarified by explicitly discussing mechanism versus function, or efficient versus final causes, rather than the ambiguous proximate versus ultimate. Part of the problem is disagreement about what constitutes a cause. Many would restrict “cause” to efficient causation and would deny that functional explanations are causal. Tinbergen (1968) belonged to this semantic camp. With respect to his four questions (survival value, mechanism, development, evolution), he wrote:
The first question, that of survival value, has to do with the effects of behavior; the other three are, each on a different time scale, concerned with its causes. (1412, emphases added)
However, a distinction between how (incorporating how come)? and what for? remains useful. Questions of mechanism and adaptive function invite different kinds of answers. Intentional language is often the most natural way to talk about adaptations. Accusations that such language illegitimately invokes supernatural entities or conscious agents are usually no more than willful misunderstanding and petty point-scoring.
In retrospect, Mayr’s choice of “ultimate cause” to describe answers to how come? was unfortunate, because ultimate cause has an “affinity” for the finalistic what for? The reason is partly etymological. Ultimate refers to something “at the end” and final causes explain a thing by an “end.” “Ultimate cause” and “final cause” are therefore easily perceived as synonyms. The reason is partly historical. The appearance of plan and purpose in Nature was commonly ascribed to God as the final or ultimate cause. Once appearances of purpose were explained as artifices of a blind (rather than divine) watchmaker, “ultimate cause” was readily interpreted as adaptive function. Mayr exacerbated the ambiguity by associating ultimate cause with why and proximate cause with how, a contrast that had historically been associated with the distinction between efficient causes how? and final causes why?
Modern defenders of a proximate–ultimate distinction interpret ultimate causes as answers to the teleological what for? They want recognition that “what is it for?” and “what is the mechanism?” are different questions with different kinds of answer and that both questions lie within the domain of science. They want respect for their interest in adaptive explanations. Modern critics of the proximate–ultimate distinction interpret ultimate causes as answers to the historical how come? They want recognition that answers to “what are the mechanisms?” are important for understanding “how did it evolve?” They want respect for their interest in the role of developmental mechanisms in evolutionary processes. Defenders and critics are talking past each other and, if they took a deep breath, would probably concede each other’s principal concerns.
“Proximate cause” does not generate much fervor and remains useful, whereas “ultimate cause” arouses passions and means different things to the contending parties. Mayr (1961) explicitly disavowed what for? in favor of how come? Therefore, critics of the proximate–ultimate distinction can plausibly argue that their interpretation of ultimate cause is closer to Mayr’s (1961) intent, whereas defenders of the distinction interpret the difference between ultimate and proximate causes as corresponding to the why versus how contrast of Mayr (1974). I recommend that we follow Mayr’s (1993) lead and use “evolutionary cause” to refer to physical causes operating in evolutionary time, not just adaptation by natural selection. Proximate causes have acted in every generation in the past and can form a legitimate part of evolutionary explanation. On the other side of the contrast, those who care about distinguishing the functional what for? from the mechanistic how? should abandon “ultimate cause” and simply talk about purpose and mechanism or employ the venerable distinction between final and efficient causes, with final causes interpreted as ascriptions of function or anticipated ends. I do not know any vitalists among my colleagues, nor any colleague who believes that divine intervention plays a role in biological explanation. The presumption should be that when evolutionary biologists use teleological language they refer to adaptation by natural selection. There is plenty to argue about without squabbling over language.