By: Anthony Aubel
In philosophy of biology, functional explanations have come to play an essential, yet controversial, role. The concept of function stemmed from teleological explanations that emphasized the purpose or goal-directed nature of phenomena and structures observed across living organisms. Teleological terms, such as ‘purpose’ or ‘goal, however, have often introduced ambiguities in biological explanations which is often a source of potential confusion since they imply a type of underlying intention or design that are not supported by scientific evidence. To some extent, however, the use of the term ‘function’ has successfully eliminated this ambiguity helping shift the emphasis to the causal roles that various biological components and systems play in living organisms.
But functional explanations have raised several new challenges in philosophy of biology. For instance, there are disagreements over clear and universally accepted definition of function and debates surrounding unified views of function. Such challenges have significant implications for understanding explanatory strategies in biological explanations. One such debate over unity is highlighted in Peter Godfrey-Smith’s paper ‘Functions: Consensus Without Unity’ where he provides a lucid argument against a unified view of functions in biology. Conversely, Philip Kitcher, in his essay ‘Function and Design’, aims to unify the concept of function by appealing to an underlying source of ‘design’. In this essay, I plan to engage in this debate and show how a non-unified view of function provides a more accurate understanding of different explanatory strategies in the philosophy of biology. I will employ the notion of Wright-Functions and Cummins-Functions coined by Godfrey-Smith to argue that in biological explanations, a dual-perspective is more advantageous than a unified view.
The unified view of function:
Godfrey-Smith begins his argument by first distinguishing between two main types of functions, what he calls ‘Wright functions’ and ‘Cummins functions’, in reference to Larry Wright’s and Robert Cummins’s conceptions of functions, respectively. Wright-functions (WFs hereafter) are defined from an etiological (i.e origin) perspective entailed by a particular component’s historical and evolutionary roles played in the survival and reproduction of an organism or a system. Specifically, a component's WFs are the ones that play a role dictated by Darwinian natural selection. Cummins functions (hereafter CFs), on the other hand, are defined as those features of an organism’s or system’s components that are currently playing a causal role in its operation and organizational structure. Unlike WFs, CFs do not figure in the historical or evolutionary origins of a component, but rather with its contribution to the overall functioning of an organism or system in its present state (Godfrey-Smith, 1999).
This distinction between WFs and CFs lays the grounds for Philip Kitcher’s attempts to argue for a unification theory of functions. In “Function and Design”, Kitcher defends the claim that “all biological attributions of function take place in a context characterized by design” (Kitcher, 1993). He uses the concept of design to construct a unified theory of function that aims to bridge the gap between WFs and CFs by presenting a middle ground between etiological and causal role explanations. This view is understood as serving an all-encompassing basis that applies to all biological components and systems where Kitcher proclaims to be necessary for a coherent and consistent interpretation of biological phenomena. For a trait to be considered designed for a function, he appeals to a concept of unity that is founded on the idea that “the function of an entity S is what S is designed to do” (Kitcher, 1993). He attempts to set forth criteria for traits where if met by both the causal (i.e. CFs) and etiological (i.e. WFs) roles, then the trait can be considered designed for a function. For example, a trait is said to meet its causal role if it contributes to an organism's survival and reproduction and if the trait has played a key role in the evolutionary history of the organism that led to its selection by conferring fitness.
The argument against unity:
Godfrey-Smith challenges Kitcher's view by pointing out two distinct sources for traits that contribute to a system’s response to the environment but do not seem to be products of selection: chance and constraint (Godfrey-Smith, 1999). Let us have a closer look at the role these two sources play in the debate.
Godfrey-Smith argues that chance-driven events that arise naturally based on random events present cases where Kitcher’s account fails to explain features that contribute to an organism’s fitness, even in cases where the traits happen to not have any clear history of selection. For instance in evolutionary biology, the term exaptation is used to explain properties of a trait that have been co-opted to serve a new function not directly related to its original function (Allen and Neal, 1996). Cases of exaptation pose a clear challenge to Kitcher’s attempts for unification of functions (what Godfrey-Smith refers to as ‘false unity’) since the traits co-opted in this way fail to meet the etiological criterion because their current function may not have played a significant role in the evolutionary history of the organism.
Kitcher’s proposal is similarly examined with respect to how selection of properties in particular organisms (i.e. fruit flies) are influenced due to some environmental constraints. For instance, mature fruit flies undergo changes in size based on varying temperatures in the environment. This property appears to be an adaptation that is related to avoidance of moisture loss. Here, Kitcher’s analysis based on design would suggest that the physiological properties causing this change have the function of adjusting the fruit flies’ size based on the environmental factor. But it would be wrong to assume these properties as functional since they are the result of constraint and hence physiologically inevitable. These observations support a non-unified view that recognizes both WFs and CFs because such functions play roles in two different explanatory modes within philosophy of biology. Attempts of merging them together would, as Godfrey-Smith points out, create a false unity that obscures important distinctions between selected effects and fortuitous benefits.
Godfrey-Smith’s argument from constraint is supported by evidence from studies in ecological developmental biology where the interplay between genetic and environmental influences are emphasized. The so called epigenetic factors are essential for understanding an organism’s function in ecological context. For example, Gilbert and Epel’s work on several organisms show clear examples of epigenetic phenomena. During their development, tadpoles of a specific species of toads, for instance, diverge into two distinct morphologies based on constraints in their diet. Tadpoles that grow in environments abundant in protein sources (i.e. tiny fish and shrimps) which they can consume, develope to be carnivorous displaying larger jaw muscles, broader mouths, and long and convoluted guts. These characteristics can be beneficial for the organism’s fitness. In contrast, those that grow up in environments low in protein sources and must resort to eating debris and other plant-based contents, develop into omnivores showing opposite morphological characteristics (i.e. smaller jaw muscles and mouths, short and smooth guts). Such characteristics, however, can be disadvantages for the organism (Gilbert, 2012). These differences in morphology reflect the distinct causal role attributed to CFs where they make a contribution to capacities and dispositions of the tadpoles when confronted with an environmental variable. Yet the events that lead to some tadpoles developing in one environment or another is random. But in Kitcher's view, these effects will be considered functions only in the case where the effects happen to be beneficial for the organism where it helps it meet the demands of the environment. Thus, attempting to reconcile the distinct CFs with WFs in this context distorts our understanding of these systems.
To further support and illustrate the benefits of a non-unified view in understanding biological systems, let’s examine more closely the evidence from some other cases.
Consider the following specific case from human anatomy: the appendix. It has historically been considered by anatomists that the appendix is a vestigial organ with no apparent function designated to it. However, recent research suggests otherwise. The appendix may indeed serve a causal role function (i.e. CFs) as a reservoir for beneficial gut bacteria which contributes to the maintenance of gut microbiomes (Vitetta, Chen and Clarke, 2019). In this case, the selected effect view of the appendix would suggest that it bears no current adaptive function since it is a remnant of the tail-end part of the large intestine from ancestral species. Kitcher’s attempts to reconcile for both of these findings would likely fail and may not adequately account for them in a unified way.
The explanatory significance of causal role (CFs) and selected effects (WFs) functions is also found in molecular biology. Often it is observed that the same molecular components in cells display multiple functions depending on context. This context-dependence of molecular features is sometimes referred to as one-many argument in philosophy of biology literature (Griffiths, 2008). Critics claim that one-many phenomena “...is a robust empirical generalization that a molecular pathway may have different effects in different cellular contexts; the same pathway can be involved in different functions in different species or in different parts of an individual”(Gilbert and Sarkar 2000, Laubichler and Wagner 2001, Burian 2004). Empirical evidence from gene regulation studies in bacteria supports this view. For instance, the so called lac operon, which refers to a cluster of genes that function in lactose metabolism in certain bacteria (i.e. E.coli), dictate the production of enzymes while also serving as a regulatory mechanism for gene expression (Griffiths, 2008). This indicates the same molecular component can have multiple functions depending on the context which challenges the unification view in explanatory strategies.
Problems with Kitcher’s concept of design for function:
Furthermore, Kitcher’s concept of design raises metaphysical problems in biological explanations by invoking questions around intentionality imposed onto natural processes. The use of the design metaphor, for instance, may suggest that nature operates with an underlying purpose and hence begs an appeal to teleological explanations where the emphasis is placed on the purpose or goal-directed nature of phenomena and structures observed across living organisms. Such claims however are not supported by empirical science. Teleological explanations have historically introduced ambiguities in biological explanations which is often a source of potential confusion. Functional explanations, in contrast, have successfully been able to avoid any reference to purpose, design, or intentional language that can obscure descriptions in biology. Rolf Gruner, in his analysis of Teleological and Functional Explanations, for instance, provides a clear distinction between cases where a goal is pursued consciously and cases where a goal is not consciously pursued. He asserts that “the word ' functional' comes in very handy and by using it the associations with purposes, desires, intentions, etc., are eliminated” (Gruner, 1966). In light of modern science, mechanistic accounts that seek to explain complex phenomena through the interactions of their constituent parts, have increasingly overshadowed their teleological counterparts. Mechanistic explanations not only have been able to successfully incorporate functional explanations, they also have several advantages and offer detailed explanations of biological processes without appeal to a source of design or designer. Kitcher’s appeal to design can easily conflict with such naturalistic and mechanistic understanding of biological processes.
Kitcher’s design-centered perspective is also prone to circular reasoning when attempting to unify WFs and CFs since both types of functions often rely on similar concepts (i.e. fitness or adaptation). Potential issues of circularity or explanatory redundancy may arise since there is a lack of clear, independent criterion for distinguishing between the two types of functions. In the example of vertebrate heart discussed earlier, for instance, WFs would emphasize the organ’s role as a trait that has been evolutionarily selected for due to its contribution to the organism’s fitness. But CFs would examine its contribution to the overall circulatory system and the causal role it plays (i.e. distributing oxygen, nutrients throughout the body). However, circularity in explanations can arise when the heart’s design can not only be attributed to blood pumping capacity influenced by selective pressures favoring efficient circulation in organisms ancestors, but also allowing it to effectively perform its systemic function within the circulatory system. But this would conflate matters since the mere concept of design relies on the very selection pressures and adaptations that WFs and CFs seek to explain, hence leading to potential circular reasoning. On the other hand explanatory redundancy may arise where attributing the function of the heart to its design inadvertently double counts contributions afforded by both WFs and CFs. This in turn can reduce the distinct explanatory powers offered by WFs and CFs and detract from further insights into the heart’s role in organisms’s overall physiology and evolutionary history.
Advantages of non-unified (dual) view in biological explanations:
The advantages of a dual (non-unified) perspective using WFs and CFs in explanatory strategies is particularly useful in cases where a component’s historical role and its current role are not identical. Consider for instance the case of the vertebrate heart where we can clearly apply both a WF and CF. The function of pumping blood to facilitate the circulation of oxygen and nutrients throughout the body as determined by an evolutionary history perspective can be explained as WFs. Whereas CFs accommodate for these properties in terms of an organism’s current physiological needs or state by maintaining blood pressure and regulating the distribution of blood flow. We can see that recognizing this distinction between WFs and CFs in explanatory strategies allows for more comprehensive and accurate understanding of phenomena in biological explanations. This dual (non-unified) view should not be construed as an arbitrary introduction of complexity in our explanations; rather, we ought to acknowledge the inherent complexity of biological organization which requires diverse explanatory needs.
We also saw in earlier examples provided earlier that keeping WF and CF type explanations separate can help us have a better understanding of the interactions between genotypes, phenotypes, and the environment. Organisms often exhibit a wide range of developmental plasticity where their functions (and traits) may change based on environmental factors. For instance some organismal phenotypes are plastic; meaning the organism is capable of modifying certain set of its traits in response to environmental changes. The factors that influence such capacities for plasticity can be traced and studied using the WF from an evolutionary, historical perspective. Employing CFs can further help in examining the functional consequences of the adjustments that an organism is able to undertake.
The case for plants' ability to adjust the rates of photosynthetic (the molecular processes of using sunlight to make food/energy) reactions in response to the amount of light present in the surrounding environment presents a good example of such phenotypic plasticity. There are plenty of scientific evidence that highlights the evolutionary pressures that drive the development of such plastic capacities in plants. WFs, therefore, play a key role in this aspect of explanations. In addition, CFs help elucidate the functional role of this plasticity as to how it relates to, for instance, optimization of plant's energy production and growth. Therefore, this dual perspective allows biologists to apply CFs to examine the causal roles of traits while applying WFs helps in investigations of how such traits have been shaped throughout evolutionary history, and by natural selection, in response to different environmental conditions.
Furthermore, the role of genetic variability influencing biological systems, for example, can be more thoroughly investigated by using WF type explanations where the evolutionary basis of traits are emphasized. On the other hand, the study of how those genetic variabilities impact the functionality of the organisms can be more readily enabled by CF type explanations. During the industrial revolution, for instance, the peppered moth population underwent a shift in color from light to dark due to selective forces at play as a response to air pollution. Interestingly, this selective advantage towards dark phenotype, conferred better camouflage when the organisms sat on soot-covered trees (Cook and Saccheri, 2013). This effect can be understood through the lens of WF approach by revealing evolutionary pressures shaping the traits involved. Meanwhile, an explanation from CF perspective allows researchers to investigate the causal roles of the various moth phenotypes within such environments; for instance, the moth's ability to avoid predation, seek mates and reproduce are phenotypic contributions to the overall performance of the organism that can be more effectively examined by employing CF type explanation. This, in turn, can help us better understand how genetic variability in the organism's population influences their success and adaptability within different environmental contexts.
The cases and the evidence that we have explored in this essay demonstrates that a non-unified view of functions that embraces both WFs and CFs, can offer a more accurate understanding of various explanatory strategies in philosophy of biology. Through critical examination of Kitcher's unified approach and its limitations, we have attempted to illustrate the importance of maintaining a dual perspective in biological explanations by showing its effectiveness in addressing questions around biological trait's functionality and adaptive nature. Ultimately, adopting a non-unified view enables a deeper understanding of the intricate interplay between evolution, adaptations, and functions in biology.
Reference List
Allen, C. and Neal, J. (1996) ‘Teleological notions in biology’.
Boorse, C. (1976) ‘Wright on Functions’, The Philosophical Review, 85(1), p. 70-86.
Burian, R.M. (2004) ‘Molecular Epigenesis, Molecular Pleiotropy, and Molecular Gene Definitions’, History and Philosophy of the Life Sciences, 26(1), p. 59-80.
Cook, L.M. and Saccheri, I.J. (2013) ‘The peppered moth and industrial melanism: evolution of a natural selection case study’, Heredity, 110(3), pp. 207–212. doi:10.1038/hdy.2012.92.
Cummins, R. (1975) ‘Functional explanation’, Journal of Philosophy, 72, p. 741-764.
Gilbert, S.F. (2012) ‘Ecological developmental biology: environmental signals for normal animal development’, Evolution & Development, 14(1), pp. 20–28. doi:10.1111/j.1525-142x.2011.00519.x.
Gilbert, S.F. and Sarkar, S. (2000) ‘Embracing complexity: organicism for the 21st century’, Developmental dynamics: an official publication of the American Association of Anatomists, 219(1), p. 1-9.
Godfrey-Smith, P. (1999) ‘Functions: Consensus without unity’, Function, selection, and design, p. 185-197.
Griffiths, P. (2008) ‘Philosophy of biology’, Stanford encyclopedia of philosophy [Preprint].
Griffiths, P.E. (1993) ‘Functional analysis and proper functions’, The British Journal for the Philosophy of Science, 44(3), p. 409-422.
Gruner, R. (1966) ‘Teleological and Functional Explanations’, Mind, 75(300), p. 516-526.
Kitcher, P. (1993) ‘Function and design’, Midwest studies in philosophy, 18, p. 379-397.
Laubichler, M.D. and Wagner, G.P. (2001) ‘How Molecular is Molecular Developmental Biology? A Reply to Alex Rosenberg's Reductionism Redux: Computing the Embryo’, Biology & Philosophy, 16(1), pp. 53–68. doi:10.1023/a:1006750126784.
Millikan, R.G. (1989) ‘In Defense of Proper Functions’, Philosophy of Science, 56(2), pp. 288–302. doi:10.1086/289488.
Sober, E. (2014) The nature of selection: Evolutionary theory in philosophical focus. University of Chicago Press.
Vitetta, L., Chen, J. and Clarke, S. (2019) ‘The vermiform appendix: an immunological organ sustaining a microbiome inoculum’, Clinical Science, 133(1), p. 1-8.