Evolutionary Functionalism - Evolutionary Explanations of Behaviour

3. Evolutionary Explanations of Behaviour

3.2. Evolutionary Functionalism

Much of biological explanation is functionalist. Traits and processes are understood by their purpose for and within the organism, and this pur-pose is naturalized in one way or another. One such way is evolutionary functionality. This can also be a perspective on social behaviour, includ-ing human behaviour. If this is the chosen perspective, questions arise.

Functionality of what? For what? Functionality in what sense? Some of these questions will lead to the questions about the relation between in-dividual and above-inin-dividual level descriptions and explanations.

3.2.1. Adaptivity and the “Consensus without Unity”

One way to characterize evolutionary functionality – or even to give a general explication for the biological notion of function – is to say that a trait can be explained as a component in a mechanism that produces

something that is useful for the organism. This participatory role in a mechanism is the trait’s function. There are different ways to under-stand what exactly this means, and there is not even a clear view of what the analysis of function is supposed to achieve. Functions seem to be useful in distinguishing between purposeful and accidental (for example, the heart pumping blood versus making noise), they are supposed to be explanatory (why we have a heart), and they have a normative aspect, that is, distinguishing dysfunctional (not pumping properly) from functional, and malfunctioning (ventricular fibrilla-tion) from functioning, and all these dimensions of functionality are contested (Garson 2016). Pluralism about function concepts is widely accepted (see Godfrey-Smith 1993; Amundson & Lauder 1994;

Griffiths 2006; Bouchard 2013; Garson 2016). As Peter Godfrey-Smith (1993) put it, there is a “consensus without unity.” I will follow this consensus and, rather than defending a view of functions, I will pre-sent the “basic concepts” of function and discuss them in connection with our current purposes. The three concepts I consider to be the basic concepts of function are the etiological function, the causal-role function, and the adaptive function.

The etiological function of a trait (Wright 1973 & 1976; Millikan 1984; Neander 1991) is whatever the trait does that explains its exist-ence. Larry Wright (1973: 161) defines it thus:

“The function of X is Z means: a) X is there because it does Z and b) Z is a consequence (or result) of X's being there.”

The reason for including X doing Z as the cause for the existence of X is to naturalize the apparent purposefulness and explain it as well as to provide criteria for distinguishing the “real” function from every-thing else that results from X. There is no theoretical necessity to re-strict the use of the concept of etiological function to the evolutionary dimension of biological causes, but in practice, and especially within the evolutionary perspective, a biological trait has etiological function only if it is an adaptation. Ruth Millikan (1984) and Karen Neander

(1991), for example, defend the selected effect variant of etiological function. In this view, the function is whatever it does so that the or-ganism becomes fitter, which in turn has caused the trait to be selected.

Whatever else it does for the good of the organism is just a set of other consequences. This is also a way to understand the adaptationism de-bate. Empirical adapationism is a claim that (most) traits have an etio-logical function, and both explanatory and methodoetio-logical adaptation-ism give a trait’s etiology a special role in describing the trait. If this were the only way to understand functions, many biological traits, including some vital ones, would not have functions at all: useful by-products and exaptations could not be considered to have functions.

The causal-role function (Cummins 1975 & 1983) is an articula-tion for a funcarticula-tion that is used to analyse the trait from the point of view of its causal role in the system that it is a part of, from the point of view of the functioning of the whole system. It is a more liberal ac-count that does not require the function to contribute to its own exist-ence. The proponents of this approach argue that there is a need for a concept of function like this, especially in physiology, neuroscience, ecology, and ethology (see Amundson & Lauder 1993; Bouchard 2013;

Craver 2013).⁶¹ The causal role function of a trait cannot be whatever causal consequences it has as a part of a system. Etiological functions are too demanding for many purposes, but causal role functions re-quire a perspective from which the functions of the parts are analysed to rule out accidents and unimportant effects. Therefore, the definition needs to include a perspective (see Cummins 1975 & 1983):

An X has a function F in the system S in relation to an analytic perspective A that is adequate to the S’s capacity to G

if and only if

61 The different notions of function should not, however, be automatically taken to be differences that follow from the different fields of biology (Garson 2016).

100

X has a capacity to produce F (or have a role of F) in the system S as a part of a process resulting in G under the description within A that has a reference to an X capable of F.

Unlike with etiological functions, there is no objective way to say how to choose A and G given any S and X. An obvious perspective is what-ever explanatory interest we have (Rattcliffe 2000; Lewens 2000), which frames the contrastive-counterfactual questions we ask, and, therefore, which causal functions are constitutive parts of a mecha-nism we use for explanations. Therefore, we can say that:

An X has a function F if and only if 1) there is an effect G that

2) we are explaining from the perspective A 3) with a mechanism M

4) that has X 5) with a capacity to F 6) as a constituent part of M.

Causal role functions refer to real causal relations, capacities and pro-cesses, and are explanatory of the whole system and its capacities (they tell what the part does from the perspective of a wider system), but not all of the functions of a trait are equally interesting from the specific explanatory point of view. For this, we need some further cri-teria. An important factor is that biological organisms are functional wholes that have architectures, as discussed above. Whatever per-spective a researcher takes, this internally functional architecture and its mechanistic constitution and maintenance (that is, how the organ-ism stays alive) are a natural part of that perspective. An important dimension in the meaning of “function” is that the trait contributes to this instead of being dysfunctional (for the whole) or malfunctioning

101

(that is, causing harm to the whole organism by not operating in an orchestrated manner).⁶²

The overall architecture, however, is not only a functional whole in itself. It must be functional in its environment, too – as a whole, even if not all its characteristics are functional. The organism might not be optimal, and it might not be “well-adapted”, but it has to have the means to operate and survive in its environment. Even if the or-ganism’s design is widely sub-optimal, how it nevertheless survives in its environment is what is important in explaining the organism’s characteristics. For example, whales having lungs is not exactly an op-timal solution, and the explanation for their existence is in the contin-gencies of the evolutionary past, but understanding the characteristics of the whale lungs and the whole respiratory system must include ad-aptationist thinking – albeit acknowledging that the functionality of the system centres around something that is highly dysfunctional it-self. Much of the design of a given organism may be due to phyloge-netic inertia and the path-dependent nature of evolution, but some of this is adapted and some parts have been utilized for new functions (as exaptations). These characteristics are not (historically) adapta-tions for their use, but they are, nevertheless, a part of the organism’s functional design from the point of view of its overall functionality in its environment. The architecture needs to stay, in the course of evo-lution, an orchestrated whole, and functional enough in the environ-ment the organism lives in for it to be able to live in it, even if sub-optimally. If the biologists are interested in describing the functional

62 This is not always the case. If we want to understand the biology of a disease or impairment, it is not enough to know that something is malfunctioning from this point of view (although it is the starting point; see Wakefield 1992;

Adriaens & Andreas De Block 2011), but we might be interested in how the disease, a cancer for example, functions from the point of view of what it is doing that is harmful for the organism. Given knowledge about some part of the “mechanism” of cancer (for example, its ability to form an internal blood circulatory system), we can try to prevent it growing based on this knowledge (see for example Tammela et al 2008).

102

architecture of an organism, it is only reasonable to include an ecolog-ical perspective, which in turn is closely connected to evolutionary considerations. Furthermore, mechanisms used in biological explana-tions require an idea of what the mechanism is for, and a functional analysis is needed for this (see Craver 2013; Garson 2013).

Whatever role the trait plays in the overall adaptive functionality of the organism in the environment in which the organism lives is a natural starting point for understanding the organism. Even if the adaptive functionality from this point of view is not the result of an atomistic (historical) adaptation, an ahistorical adaptationist perspective may be a useful heuristic to decompose the organism’s functional ar-chitecture in its ecological context. Exaptations and structural charac-teristics that are phylogenetic relics can be treated as functional prop-erties of the overall design – if one steers clear of making historical claims or optimality assumptions about them. This also seems to reflect real biological practices: biologists are interested in the adaptivity of an organism and its traits in its current environment as an essential part of understanding the organism without speculating about their evolutionary history (see Amundson & Lauder 1994; Walsh 1996;

Wouters 2003; Cuthill 2005; Sherry 2005). Justin Garson (2016), how-ever, voices a suspicion that this practice may implicitly rely on adap-tationist thinking. This concern is warranted (see also Cuthill 2005), but regardless of what some biologists may presume of the connection of adaptive functionality and historical adaptation, there are other jus-tified reasons to be interested in adaptive functionality alone.

For starters, there are other reasons than a historical adaptation process that lead into ecological adaptivity, as I mentioned above.

First, there is habitat selection: some organisms can choose an environ-ment (recognizing environenviron-mental clues; this is habitat choice), or they may randomly end up living in an environment in which they are eco-logically more functional, with increasing reproduction rates. Similar adaptive optimization takes place, only without any change in the or-ganism, and an adaptationist perspective can be a useful ecological perspective even when we know that the organism has not evolved in

103

that specific environment. (Morris 2011.) Furthermore, if there are constraints on adaptation in some environment, another environment may be a better fit for the species: either it is directly better, or it is easier to adapt to within the given evolutionary constraints. A whole species may change its niche or migrate to another region – which is an adaptive process, too, producing adaptive functionality without organism-changing (Darwinian) evolution. This might happen when there is a change in environments (for example, a climate change) and a population changes its geographical location to keep its habitat. Or-ganisms may also find new environments that fit them as well as or even better than the environment they evolved in.⁶³ Furthermore, gen-eralist organisms with behavioural plasticity may be able to utilize the features of the environment that are utilizable for them (for example, as food) beyond just those exact features that the capacities were se-lected for. Developmental plasticity in turn may lead into direct adap-tation to new environments (West-Eberhard 2003). Plasticity is espe-cially important in humans. Yet another adaptive process to be taken into account is niche construction (Odling-Smee, Laland & Feldman 2003): the organism not only chooses the habitat but may have adapted to actively change its environment in ways that make the or-ganism a better fit to it, and further evolves into those environmental products, making the organisms and their environments co-evolving interactive wholes.

All the above ways of becoming adaptive without an organism-changing adaption process (or Darwinian selection) are further

63 There are, for example, species that do better in urban than natural environ-ments, the most extreme example being the rock dove (Columba livia) – the feral pigeon. They did not evolve for urban environments, but they have found an ecological fit in these environments that is better than what they find in natural environments, after which certain features in their behaviour have further adapted to urban environments (Rose, Nagel & Haag-Wackernagel 2006). It is sensible to study the pigeon ecology as adaptively functional to cities even if only the finishing touches were produced by adaptation to this environment.

104

reasons to abandon empirical and historically explanatory adpapta-tionism – and they should be abandoned. At the same time, they ad-dress further needs, beyond the Darwinian adaptation process, for adaptive functionality analyses of traits. First, a description that in-cludes a functional account of the trait’s role in the overall form of life of the organism provides a deeper understanding of the organism, and the natural basis for such an account is from the point of view of what all organisms fundamentally do: stay alive, sustain themselves, and reproduce (see Wouters 1995; 2005; 2013). How an organism does this in its environment is simply a part of its biological description.

Arno Wouters (1995) calls explanations referring to these features of the organism viability explanations: a trait is explained by how it satis-fies an organism’s needs. This is not a causal explanation for the trait’s existence, but it is a functional explanation of how the trait contributes to the viability of the organism. Wouters even proposes that it is this idea of functionality that biologists use in their practices, never an eti-ological function, unless they are doing evolutionary biology specifi-cally (Wouters 2013).

Note that “viability” only refers to the means to survive in the en-vironment, and while the notion of “survival of the fittest”⁶⁴ has become the misleading slogan for natural selection, one does not have to be the fittest to survive, and natural selection goes beyond mere survival. Alt-hough optimality is only a theoretical endpoint of the adaptationist se-lection models and cannot always be expected to be found (for the rea-sons discussed above), natural selection can be expected to result in forms somewhere in between mere survival and optimality. And the other way around: natural selection is a partial explanation for viability and everything beyond. We do not need to restrict the considerations of adaptivity to optimality; adaptivity can be seen as a range of variation

64 The phrase does not exist in the original Origins (Darwin 1859) – the origin of the term is Herbert Spencer (1864). Unfortunately, Darwin endorsed it and added it to later editions of the Origins – mostly because the expression “nat-ural selection” has other misleading connotations, namely that of there being something that literally selects something. (See Gould 2002.)

105

from survival to optimality. Third, a functionalist description from an evolutionary perspective is still explanatory in a more minimal sense that I have called “ahistorical adaptation” above. Viability and adaptiv-ity of a trait are more fundamental properties in the evolutionary expla-nations than the selection process: selection is based on the comparison between viability and adaptivity of individuals (or other units). I will expand on some of these points next, and I will discuss the third “basic concept” of function, the adaptivity function.

3.2.2. Adaptive Functionality and a Taxonomy of Functions One of the main problems of adaptationism in all its forms is the pre-supposition of atomism (that is, each trait has a unique, isolated task it is selected for), when an organism is a holistic system (Gould &

Lewontin 1979; Gould 2002). The organism’s parts simply cannot be functionally independent, atomistic adaptations. At the same time, the whole must have some degree of adaptedness. We can switch from atomism to a holistic approach and take the overall adaptedness of the organism to be the starting point of a functional analysis. We can also approach the environment as holistic surroundings with multiple simultaneous adaptive challenges with some qualitative variation. If we analyse individual traits from this perspective, we end up with a vastly different idea of what counts as functional in the first place. This overall adaptedness, in turn, can only come from some sort of adap-tation process (be it Darwinian selection, habitat selection, or some-thing else) that is at least partly guided by the logic of natural selec-tion. This is the case even if the parts of the adaptive whole have not evolved for the adaptive function they have now, and even if the adaptedness is not optimal. Moreover, even if a trait has evolved be-cause of some other evolutionary factors, natural selection may some-times be a part of an explanation of why it remains.

It seems therefore that ahistorical adaptation perspective has two explanatory functions. First, even if natural selection is historically only a partial explanation for the apparently purposeful design of the

106

whole organism, if one is interested in an organism’s way of life in its natural environment, then the point of view of adaptedness (or func-tionality in environment) has an explanatory function. Second, ahis-torical adaptation perspective has counterfactual power. If a change took place in the environment for which the organism is now fit, either there would be some change in the organism or its geographical loca-tion, or it would be doing worse than in the current environment, pos-sibly facing the threat of extinction. There might be ecological conse-quences that loop back to the functionality of the organism in the changed environment. In this view, a particular trait needs not to have evolved for whatever its role is from the point of view of this adaptive whole; it is sufficient just to play a role.

This approach is not only a heuristic approach to the organism’s design, but also minimally explanatory: it identifies a selection pres-sure and a corresponding trait and has, therefore, explanatory power over a set of what-if-questions about possible alternative traits relative to which it would be selected for or against, as articulated above. The components of the natural selection mechanism with the trait’s adap-tive value included are in place, even if they were not included in the evolutionary historical explanation of the trait. Even if natural selec-tion did not guide an evoluselec-tionary process in the past, it is an effective cause in the on-going process of evolution, and it is predictive of the future. To make an even stronger claim, this is the case even if devel-opmental constraints prevent the relevant alternatives, and these con-straints, instead of selection, are the cause of the persistence of the trait. This is because the selection pressures exist even then, too. This supports the counterfactual that if the developmental constraints broke, selection would still maintain the trait, selecting against the

In document Evolving in Groups : Individualism and Holism in Evolutionary Explanation of Human Social Behaviour (sivua 109-140)