However, I am confident that the distinction can be rigorously made in a [Page 1255] high percentage of cases, probably a firm majority. If direct knowledge of historical sequences from paleontological data established the only path to resolution, then imperfections of the fossil record would preclude resolution at sufficient relative frequency. But evolutionary biologists can also reach firm conclusions about historical sequences from cladistic reconstructions of phyletic topography based upon the distribution of traits among living organisms. (I did not participate in the cladistic revolution within systematics, and I always maintained a cautiously critical, if basically supportive, attitude towards this scientific reform. But, and now in retrospect, I must credit cladistics with the signal achievement of devising a workable methodology for inferring historical and genealogical pathways from the distribution of features among modern organisms — thereby making the reconstruction of biological history operational as a generality, and not only in special cases of adequate evidence from the fossil record. Such an accomplishment marks a fundamental advance for a historical science like evolutionary biology — and cladistics must therefore be celebrated, if for no other reason, as one of the great achievements in the history of evolutionary thought.) In any case, my previous discussion of Arnold's work on sand-diving and crevice-dwelling lizards (pp. 1234–1238) illustrated the use of both criteria (direct historical sequencing and cladistic reconstruction) in distinguishing adaptation from exaptation in biological data.
The spandrels of San Marco provide an even clearer case — relatively free from biasing preferences imposed by our engrained assumptions about biological structures — for the testability, and hence the operationality, of distinctions between adaptation and exaptation, using both major criteria of direct data and inferences from taxonomic structure. In fact, my initial choice of this example stemmed explicitly from the availability of definitive data in the rarely available category of preserved historical records of actual genealogical sequences.
To state the conceptual problem: One might strongly favor the hypothesis, based on structural arguments alone, that the spandrels originated as non-adaptive side consequences, and only later achieved utility in housing the evangelists — thus identifying this derived function as exaptive rather than adaptive. But the static evidence of architectural form cannot decide the historical issue, for the alternative interpretation remains logically unimpeachable, however unlikely — and resolution therefore requires evidence about actual historical sequences. That is, one might reverse the flow of causality and argue: why must I regard the spandrels as primary nonadaptations constraining a later choice of aptive ornamentation? Perhaps the four evangelists represent a primary impetus rather than a secondary accommodation. Perhaps the architect chose to build his church with domes mounted on sets of four rounded arches because he had such a terrific idea for festooning the resulting spandrels of the central dome with mosaics of the four evangelists and the four rivers of Eden. In this case, the spandrels would exist to house the evangelists, and the mosaic designs would become primary adaptations. [Page 1256]
I chose the San Marco spandrels because, in this case, we have firm evidence in both categories to reject this alternative possibility, to affirm the origin of spandrels as nonadaptive side consequences of a larger architectural decision, and to understand the mosaic designs as secondary adaptations within a space exapted for their utility.
(1) Direct historical data. We know that the spandrels were not built to house the evangelists because San Marco stood and operated in its appointed role as a church for at least three centuries before the mosaicists applied their astonishing work to a series of constrained and previously unornamented spaces (see Demus's classic four volume monograph (1984) on the history, architecture and iconography of San Marco).
(2) Inferences from taxonomic structure. Human buildings cannot be ordered as branching genealogical systems following both Linnaean and Darwinian logic — so we cannot base our inferences upon true cladograms in this case. But a taxonomy of ecclesiastical, and other large public, buildings does permit us to invoke a series of standard arguments, long utilized in the comparative anatomy of organisms as well, and all indicating both a nonadaptive original status for the spandrels, and a secondarily adaptive role for the evangelists as good designs in exapted spaces:
1. Ubiquity vs. occasionality as evidence for both priority and necessity. Thousands of Western buildings feature domes atop rounded arches — and every single one of them generates tapering triangular spaces at the intersections. These spandrels are ornamented in a wide variety of ways, each appropriate to the local circumstance. Many carry no ornaments at all (indicating that spandrels must be generated but need not bear “adaptive” designs).
2. Constancy of form vs. variety in usage. How could such a diversity of employment always generate the same housing? This particular distribution of anatomical “features” indicates that a constant form preceded the various ornaments thereof (both historically in phyletic time in some cases, and structurally in ontogenetic and architectural sequence in all cases). I have, for example, noted various religious foursomes in the spandrels of other churches — the four major Old Testament prophets, Isaiah, Jeremiah, Ezekiel, and Daniel in many cases; or, in San Ignazio in Rome (and “politically correct” by current standards of gender equality), four Old Testament heroes and their weapons: David with his sling, Judith with her sword (to behead Holofernes), Samson with his jawbone, and Jael with her tentpost (to transfix Sisera through the head). I also have noted secular themes in civic or scientific building — the four continents of Africa, Europe, Asia, and America under the main dome of the Victor Emanuel arcade in Milan; four classical lawmakers (Justinian, Pericles, Solon, and Cicero) under the glass dome in the Victorian courtroom of the Landmark Center, St. Paul, Minnesota; four mainstays of civilization (peace, justice, industry, and agriculture) in the County Arcade of Leeds, England, built in 1900; or the four Greek elements in the pendentives under the main dome at the headquarters of the National Academy of Sciences in Washington, DC! [Page 1257]
3. Suboptimal or ill-fitting designs as evidence of historical sequencing. As emphasized in Chapter 2, in my analysis of Darwin's rich and subtle methodology for historical inference, perfection covers the tracks of history, whereas oddities and imperfections often reveal both the direction and the stages of temporal sequences. When four evangelists establish an optimal design in four spandrels, we cannot determine the sequence of events from the structure alone: either the spaces came first, and the evangelists fit in later, or plans to depict the evangelists came first, and architects then fashioned the spaces as appropriate housing. But a peculiar, ill-fitting, or suboptimal design might suggest an order of historical precedence. If three spandrels housed elegant mosaics of Genesis, Exodus, and Leviticus, while similar designs for Numbers and Deuteronomy appeared all scrunched together in an ugly and overcrowded fourth spandrel, we might assume (at least as a preferred hypothesis for further test) that the four spandrels originated for a different and prior reason, and that a later mosaicist miscalculated badly in formulating plans for placing symbols of the Torah into these preexisting spaces.
Similarly, the foursomes in several sets of spandrels in European churches seem rather forced or even ill fitting, thus indicating that a fixed number of spaces (and their form) preceded any decision about embellishment. In the 16th century church of San Fedele in Milan, for example, four concepts, personified as women, decorate the spandrels under the central dome — the famous biblical trio of faith, hope, and charity (1 Corinthians, 13), with the remaining fourth spandrel occupied by religion. Three spandrels might have carried the intended design better, but architectural constraints dictated a quartet, so the designers had to draft a fourth participant, however unsanctioned by a very famous quotation. By etymology, “religion” may mean, “tying together,” but this particular woman seems more out of place than
integrative at San Fedele. Interestingly, the design of the great Romanesque pulpit in the Duomo of Pisa (the building adjacent to the structurally inadaptive but touristically highly exaptive Torre Pendente, or Leaning Tower) imposes no architectural necessity for a quartet of spandrels. Its lectern rests upon a tripartite column that expands to three ornamental spaces at the top. The three spaces carry heads representing faith, hope, and charity — all by their proper selves this time, with no fourth interloper to complete the occupation of a preexisting architectural constraint.
4. Invariable correlation of a specific form under discussion with broader structural features of the totality. As an unprofound point in this case, the number of spandrels always correlates perfectly with the number of arches supporting an overlying dome — thus identifying the spandrels as automatic side consequences of a broader architectural decision. We might question this argument if all buildings mounted their domes on four arches, and therefore always generated pendentives in sets of four. But, although four remains by far the most common number, the comparative anatomy of large public buildings includes some variety. The central dome of St. Paul's Cathedral in London rests upon eight arches, and the resulting eight spandrels feature the [Page 1258] four evangelists at the eastern end (where the sun rises upon novelty), contrasted with the four great Old Testament prophets to the west (where the sun sets on ancient ways).
Three major reasons for the centrality of spandrels, and therefore
of nonadaptation, in evolutionary theory
I therefore find the concept of spandrels, or features of nonadaptive origin as structural byproducts or side consequences of other architectural decisions, to be both coherently definable and eminently testable. The importance of spandrels in evolutionary biology must then rest on two further attributes: (1) their engagement with conventional theory in a challenging way that suggests potentially important changes or expansions in our general understanding of evolution; and (2), their establishment as sufficiently common to constitute a high percentage of biologically and evolutionarily relevant traits of organisms and other biological individuals.
I think that spandrels pass the first test in a robust manner, for their existence at high relative frequency (the claim of the second test) would challenge a key procedure of the adaptationist program that has long served as the day-to-day working methodology of Darwinian biologists engaged in the explanation of particulars. At the most basic level, we simply cannot gain an adequate evolutionary explanation for a trait by elucidating, however elegantly, however experimentally, and however quantitatively, its contribution to the fitness of the organisms or populations in which it now resides. Purely adaptationist analysis therefore cannot resolve history for two major reasons:
1. Through the principle of quirky functional shift, and Nietzsche's discordance between reasons for current utility and sources of historical origin, our understanding of how a current trait works cannot elucidate its mode of origin — an ineluctably and logically central task of evolutionary explanation, and one of the most interesting questions that any historical science can pose.
2. Adding insult to injury, even the most sophisticated documentation of adaptive value in a current feature gives us no right to assert similar adaptational control over its past states — even admitting the principle of quirky functional shift, and the possibility of strikingly different past usages, with current functions emerging as exaptations. Rather, the principle of spandrels suggests that a high percentage of traits now contributing in important ways to fitness arose for no adaptive reason at all, but rather as automatic side consequences of other forces (usually selection on other aspects of the organism to be sure, but with no direct selection on the trait under study). The adaptationist program cannot provide a full accounting of evolutionary change if a high percentage of traits originated as nonadaptive spandrels.
We must then pose the second question about the relative frequency of spandrels. If rare, they remain conceptually interesting, but minor in actual importance for the evolutionary understanding of particular lineages — the bread and butter of daily practice in our science. My broader case for the high frequency, indeed for the near ubiquity, of spandrels occupies the last section [Page 1259] of this chapter, but I raise three points here to set a framework of plausibility for frequencies far too high to ignore.
RELATIONSHIPS TO GEOMETRY AND ARCHITECTURE. As Geoffrey and other formalist thinkers recognized from the inception of evolutionary studies in biology, organisms are integrated entities, not hodge-podges of independent attributes each dedicated to a separate function. For two major reasons, this evident and venerable notion implies a great importance and high relative frequency for spandrels. First, any change in one part of the body must propagate correlated alterations to other parts. Selection may generate the original change for adaptive reasons, but many automatic consequences will probably be spandrels. Second, any adaptive feature of one organ will also express inherent and ineluctable attributes that must rank as spandrels. Most of these sequelae, although surely more numerous than adaptive aspects of the same feature, will probably remain forever irrelevant to evolutionary success of the lineage. (Bones are made of calcite and apatite for adaptive reasons, but bones are also white because the chemistry of these compounds so dictates. Invisible during life, this spandrel property of whiteness will probably never influence the evolutionary history of the surrounding organism. But evolution can also generate surprises in the same domain. Prior to the evolution of eyes, who would have predicted that the optical transparency of several enzymes and proteins might one day become relevant to their suitability for cooptation as lens crystallins?)
Thus, even the simplest and universal geometries of filling space must generate a host of spandrels to accompany any basically adaptive style of growth or biomechanical form. Moreover, by using criteria of direct historical records (in infrequent but best cases) or inference of genealogical order from cladistic reconstructions based on living species (a strong, if indirect, mode of argument almost always potentially available), we should be able to divide the useful features of organisms into direct adaptations, coopted adaptations with different original uses, and coopted spandrels — with the last category embodying a crucial challenge to strict adaptationist thinking: currently useful features with nonadaptive origins. Consider a simple example where the geometric nature of the spandrel can easily be defined, and where we may infer nonadaptive origin from the evidence of cladograms. (Examples of this kind could probably be multiplied indefinitely, and for any organism, were biologists more inclined to grant the subject more attention and explicit study):
All snails that grow by coiling a tube around an axis must generate a cylindrical space, called an umbilicus, along the axis. The umbilicus may be narrow and entirely filled with calcite (then called a columella). But the space often remains open, especially in land snails. A few species use the open umbilicus as a brooding chamber to protect their eggs (Lindberg and Dobberten, 1981).
Is the umbilical brooding chamber a coopted spandrel — a space that arose as a nonadaptive, geometric byproduct of winding a tube around an axis? Or [Page 1260] did snails initially evolve their spiral coiling as part of an actively selected design centered upon the direct advantages of protected eggs in a cigar-shaped central space? We cannot use the first method of data from actual historical sequences to resolve this question because we do not know whether the first coiled snails brooded their eggs in an umbilical chamber. But the second method of cladistics and comparative anatomy seems decisive in this case, however inferential: the cladogram of gastropods includes thousands of species, all with umbilical spaces (often filled as a solid columella and therefore unavailable for brooding) but only a very few with umbilical brooding. Moreover, the umbilical brooders occupy only a few tips on distinct and late-arising twigs of the cladogram, not a central position near the root of the tree. We must therefore conclude — both from geomet
ric logic (ineluctable production of the umbilicus, given coiling of the shell) and from the distribution of umbilical brooding on the cladogram — that the umbilical space arose as a spandrel and then became coopted for later utility in a few lines of brooders.
In an equally evident example of an automatic side consequence generated as a geometrical necessity, Megaloceros giganteus, the so-called “Irish Elk,” elongated the neural spines of its shoulder vertebrae for an immediately adaptive reason that seems well documented in fossil evidence, and rigorously validated by biomechanical analysis (Lister, 1994). These deer grew the largest antlers in the history of the Cervidae (up to 35 kilograms in weight on a 2 kg skull). Many big herbivorous mammals with heavy heads extend their neural spines to provide an increased area of attachment for enlarged ligamentum nuchae muscles that hold up the head and neck — a problem that probably affected Megaloceros more than any other deer.
The elongated spines are clearly adaptive (for internal insertion of larger muscles), but the outward expression of these enlarged bones — a raised area at the shoulders, covered with hair — probably existed as a nonadaptive spandrel at its initial phyletic appearance, an inevitable consequence of the geometry of physical space. And so this feature would have remained, as the vast majority of spandrels do, until the species's demise — except for a coopted utility that converted the original spandrel into an exaptation in this case. This raised area — literally a patch of skin spanning a space enlarged for internal reasons — became enhanced, altered in shape to a more prominent and localized hump, and festooned with distinctive stripes and colors, all (presumably) for coopted utility in mating display. I freely admit that the exaptive potential of such simple bumps and spaces must be limited, and may never exceed the primary adaptation (which originally engendered the feature in question as a spandrel) in evolutionary importance. (I also confess that I love this example largely for humanistic reasons (see Gould, 1996b). Fatty humps and coat colors do not fossilize, and Megaloceros became extinct many thousands of years ago. We only know about the hump and its colors because our Cro-Magnon ancestors painted a few of these animals on French and Spanish cave walls — see Figure 11-10.)
The Structure of Evolutionary Theory Page 200
