Return to article page

To print: Select File and then Print from your browser's menu.
This story was printed from FindArticles.com, located at http://www.findarticles.com.

Natural History
July, 2000

The Narthex of San Marco and the Pangenetic Paradigm.

Author/s: Stephen Jay Gould

Either the world was built up step by step or it unfolded according to a predetermined design--or both.

I do realize that the biggest of all bosses labored with maximal sweat and diligence during those first six days. So "perhaps it would be wise not to carp or criticise," as the Sergeant of Police once remarked in a different context (in Gilbert and Sullivan's Pirates of Penzance). Still, I must confess that I've always been puzzled by the relative paltriness of accomplishment on the second day, for the import of this episode seems almost derisory compared with the scope of all the others: light created on the first day, in the grandest of all opening acts (and especially necessary for noticing any subsequent event); earthly land, waters, and even plants in a foretaste of life on the third day; Sun and Moon, with stars as a mere afterthought, to populate the heavens on the fourth day; animal denizens of the sea and air on the fifth day; and inhabitants of the land, culminating in our exalted selves, on the sixth day.

But on this second day of relative nodding, God limited his efforts to installing a plane of division (called the firmament in the King James Bible and the sky in many modern translations, but closer to a thin metal plate in the original Hebrew), simply to distinguish the water above the plane from the water below. Big deal. Compared with all the makings and shakings of the other days, this second effort created nothing new but only constructed an artificial division within an existing homogeneity. Did God need a breather right after his initial effort in the creation business? Did he have to pause after the first step in order to recoup his courage for pushing through to the end?

The light dawned on my pugnacious ignorance (thus validating the product of the first day) several years ago, when I studied the thirteenth-century mosaics of this creation story (Genesis 1, of course) in the south dome of the narthex (the covered western porch just in front of the main entrance) of the great cathedral of San Marco in Venice--thus explaining the first part of my cryptic title (I shall soon shed light on the second and even more cryptic part, but all in good time, as the Wicked Witch of the West said, also in a different context).

In this wonderfully animistic set of scenes, radiating downward in three circles from the apex of the dome, each episode of creation features the youthful, beardless God of Byzantine tradition doing his appointed tasks for the day, while an appropriate number of angels either help out or look on (Otto Demus's authoritative 1984 four-volume work on the San Marco mosaics argues convincingly that the scenes of the dome derive from an illuminated fifth-century Greek manuscript of the book of Genesis).

In the scene for the second day--and begging an authorial pass for my irreverent anachronism in explication--God goes bowling to divide the waters by rolling a globe horizontally through the middle of the homogeneous mass, thus carving a barrier to define "above" and "below" (see detail of the mosaic on page 30). The two angels of the day stand awkwardly at the right of the scene, more in the way (for the waters remain undivided to their right) than as auxiliaries in this case. (Demus explains that the mosaic has been heavily restored and that the separation may originally have extended all the way to the right edge.)

But I understood my error only when I backed up a scene to contemplate the work of the first day. God, at the left edge (see below), creates light by making a vertical division this time--with a dark globe representing night at the right edge and the light globe of day closer to him on the left. The single angel of the first day, in a lovely touch, spreads one luminescent wing in the realm of light above his right arm (the angel faces us, so his right arm lies in the left domain of light) and one dark wing in the nighttime of his left side. Six rays, representing the work of the six days of creation, emanate from both globes.

And now I grasped the maximal depth of my error. I had demoted the second day because I had failed to appreciate the controlling theme of the entire story! I had always regarded the narrative of Genesis 1 as a creation myth based on the theme of sequential addition, as I think most people do, given our current cultural preferences for viewing history (at least its technological achievements) as a tale of accreting progress, and as my childhood instructors, both secular and religious, had certainly taught: light on day 1, something about water on day 2, land and plants on day 3, Sun and Moon on day 4, birds and fishes on day 5, mammals and humans on day 6.

But the mosaics in the San Marco narthex clearly expressed a quite different organizing theme in their charmingly naive iconography. These mosaics, at least for the first three days, tell a story about the successive separation and precipitation Of concrete items from an initial inchoate mass that must have contained, right from the beginning, all the seeds or prototypes for later realizations--in other words, a tale of progressive differentiation from unformed potential rather than successive addition, piece by novel piece, in a sequence of increasing excellence.

The striking device of drawing these divisions as alternating vertical and horizontal planes illustrates the case in a wonderfully direct way--thus emplacing, in a million mosaic tesserae, the organizing theme that, I feel confident, the author of the text intended and that artistic translators understood and expressed for at least a millennium or two before changes in Western culture blurred this context and subtly led us to read the story in a very different manner.

The tale of creation in San Marco's narthex begins in the undifferentiated chaos of full potential as, in the first scene, a dove (representing the spirit of God) flies above homogeneous waves of inchoate stuff. Then the divisions begin: a vertical plane on the first day to separate light from darkness; a horizontal plane (the bowling ball through the waters) to separate rain from rivers on the second day; and, in confirming the theme of separation rather than addition, both a horizontal and a vertical strip of land on the third day (see page 32), as God differentiates the earth below the firmament into its primary components of land and sea.

With this alternative theme of separation and coagulation in mind, I could finally understand two aspects of the story that had long puzzled me when I erroneously regarded Genesis 1 as a myth about creation by progressive addition imbued with increasing excellence. First (and to resolve my opening conundrum), the second day no longer seems anomalous, becoming instead, under the model of differentiation, an eminently worthy and weighty episode in a coherent tale. Most prescientific conceptions of the the basic elements regarded water (along with earth, fire, and air, at least in the classical Greek formulation) as one of the few fundamental principles of cosmic construction. Our ancestors did not understand that ocean surfaces evaporate to form clouds that return water as rain. Thus for them, the location of life-giving water in two maximally separated realms--at their feet in rivers and seas, and from the upper reaches of the sky as rain--must have generated a major puzzle demanding resolution at the fundamental level of creation itself. The work of the second day achieves full centrality and importance once we recognize the story of Genesis 1 as a tale about differentiation and once we acknowledge water as both a primary element and a source of all sustenance. The separation of life-giving waters into two great reservoirs, located at opposite poles in the geography of existence, certainly merits a full day of God's creative effort.

Second, the substitution of differentiation for addition also suggests that the author of Genesis 1 probably conceived of the nature of plants in a radically different manner from our understanding today. We make a primary distinction between organic and inorganic--with plants unambiguously in the first category, although usually relegated to the bottom of a rising sequence of plant, animal, and human. (Vernacular language sometimes even restricts "animal" to mammals alone, as in a kiddie card game called "Bird, Fish, Animal," which I, as a professional biologist, simply couldn't fathom until I thumbed through the deck and realized that "animal" meant only lions and tigers and bears, oh my!--and then the name of the game just drove me nuts. I shall die content if I ever persuade the manufacturers to rechristen this otherwise harmless and even instructive item as "Bird, Fish, Mammal.") Our conventional taxonomy of organic and inorganic reinforces the false view that Genesis 1 should be read as a tale of addition, for the primitive plants of day 3 should originate before any more advanced animal on day 5. But why, then, does the intervening day 4 neglect organisms altogether?

However, when we reinterpret Genesis 1 as a differentiation myth, an alternative taxonomy best explains the discontinuity between the appearance of plants on day 3 and the creation of a sequence of animals that begins on day 5 and then continues, with no further break, right to the end of the story. Day 3, with both vertical and horizontal divisions in San Marco's narthex, marks the episode of differentiation for the earth's physical potential: the division of the primal chaos below the firmament into its major components of water and land. The subsequent origin of plants on the same day (and in the next verse) suggests to me that the author of Genesis 1 viewed plants as the culminating aspect of the land's differentiation and not as a later addition from a separate organic realm, merely rooted within the substrate of another category of material stuff. In short, I would bet that the taxonomy of Genesis 1 intends to rank plants with earth, and not with animals.

In advocating the importance of ancient creation myths for our modern understanding of natural history (and hence as an appropriate subject for this magazine), I make no argument about empirical truth. Rather, I think that our primal myths teach us something important about the limits and capacities of the human mind for organizing complex material into sensible stories.

All cultures must generate creation myths; how else can we infuse order into the buzzing and blooming confusion of nature's surrounding diversity and complexity? Anthropologists, ethnographers, and folklorists have long noted the striking similarities among creation stories devised by people living in distant lands and without any known contact.

Two explanations for these similarities have generally been offered: (1) perhaps the story really arose only once and then passed from one culture to another by more efficient routes of sharing than anthropologists have recognized, or (2) perhaps the stories arose in true independence but with striking similarity guaranteed by inherited mental preferences and images--the archetypes of Jungian psychology--deeply and innately embedded in the evolutionary construction of the human brain.

But I think that we should add a third possibility, invoking logical limits to the structure of stories rather than explicitly similar content derived either by direct transfer or evoked from common residence in all human brains. (Of course, such logical limits also represent quirks of our own mentality, but this principle invokes a different and much more general aspect of consciousness than the specific images of Jungian archetypes.) After all, comprehensive stories can only "go" in a certain number of ways, and the compendium of independently derived creation myths includes so many stories that, inevitably, several tales must fall under each of the few conceivable rubrics.

These rubrics, although far fewer than the stories they must organize, still span a fairly ample range. For example, creation stories may be primarily eliminative, as when a promiscuous creator begins by populating the cosmos with all conceivable forms and then grants to nature the task of weeding out the malformed and the nonoperational. Or creation stories may be cyclical, as when new generations arise to replace their perfectly adequate ancestors and to enjoy their own transient ascendancy--so the gods of Jupiter's generation succeed their Saturnian ancestors, while Wotan and his cohort perish in Wagner's Gotterdammerung as a new day dawns at the end of four very long operas.

But only two basic frameworks of explanation can be invoked when a culture chooses to organize a creation myth as a successive and progressive series--surely a common, if not the preferred, theme among most human tales about the origin of the natural world (for whatever set of complex reasons, involving both mental preferences and natural appearances). Such tales of sequential improvement may invoke either successive addition (first make this, then add this at a higher level, and so forth) or refining differentiation (start with a big soup containing all eventual products as unformed potential and then separate/coagulate/harden, separate/coagulate/harden, and so forth). Perhaps other alternatives exist, and creation myths certainly can (and usually do) include aspects of both primal tales. But addition and differentiation surely define the primary mental territory of creation myths constructed under the guiding principle of sequential progression.

I now reach the point of necessary confession for my cryptic and self-indulgent rifle. You have probably excused me for the narthex part, already explained above. But the term "pangenetic" and the resulting full title, requires an abject plea for your indulgence. In using "pangenetic," first of all (and legitimately, for this aspect alone), I honor my hero and the inspiration for this entire series of essays: Charles Darwin. In his longest book, The Variation of Animals and Plants Under Domestication (1868), Darwin proposed, as a "provisional hypothesis" in his own words and judgment, a theory of heredity that he called pangenesis. History forgot this incorrect theory, both rightly and entirely--although curiously and by a complex route, our most salient modern word, "gene," explicitly honors Darwin's failed effort.

According to pangenesis, each organ of the body casts off tiny particles called gemmules. These gemmules circulate throughout the body, and each sex cell eventually accumulates a full set. Thus, the fertilized egg contains a complete array of determinants for growing all parts of the adult body--and embryology becomes a process of making these gemmules manifest and letting them grow. In other words, "pangenesis" expresses a pure theory of differentiation, rather than addition, for the explanation of organic development. The initial fertilized cell, like the primal chaos of Genesis 1, includes all components of the complex adult, but in unexpressed and inchoate form. Embryology then unfolds as the realization of an initially unformed but completely self-contained potential. Thus, the pangenetic paradigm, honoring Darwin's version of a larger theme, encompasses a class of models based on differentiation rather than addition, for the generation of progressive complexity in a temporal series.

Now for the self-indulgent part: The most widely cited technical paper I ever wrote (with the exception of my first article on punctuated equilibrium, coauthored with Niles Eldredge of this Museum) bore the title "The Spandrels of San Marco and the Panglossian Paradigm" (written with my colleague Dick Lewontin, the smartest man I have ever known). For a variety of reasons happily beyond (and truly irrelevant to) the subject of this essay, the spandrels paper has been unmercifully attacked by a substantial group of biologists (and also appreciated, I trust, by an even more substantial group) committed to the strictly adaptationist account of evolution that Lewontin and I questioned. Among the many published attacks, several have parodied our original title--as in "the scandals of San Marco" and even "the spaniels of St. Marx." So I thought I'd indulge myself, after all this tsuris, by writing a paper with my own title parody, albeit on a quite different subject. Sorry, folks, but at least I have laid down all my cams, and I now bare my throat.

I have little doubt that the first three days of Genesis 1 should be read as a tale of differentiation rather than addition. I also tend to view the fourth day as a continuation in the same mode--that is, God differentiates the earth below into water and land (with plants) on the third day and then, on the fourth day, differentiates the light of the sky above into Sun, Moon, and stars. But I must confess mixed feelings and signals about the fifth and sixth days. Does God now switch modes to utilize the alternative theme of addition in populating a cosmos (prepared by differentiation) with living creatures? Or does the origin of animals continue the theme of differentiation--as the air and water precipitate their living counterparts on day 5, while the land generates its appropriate forms of life on day 6?

San Marco's depiction of Adam's creation might be read as support for a continuing theme of differentiation right to the end of the story. Adam arises, dark as the earth, from the substrate of his origin. (He is not imposed upon the earth as a separate creation from astral realms. A familiar Latin pun--homo ex humo [man from the earth]--expresses the same thought, as does the familiar injunction "for dust thou art, and unto dust shalt thou return.")

My own, utterly nonscholarly intuitions lead me to view the first four days as pure differentiations: light divided from darkness on day 1, the upper water of rain from the lower water of rivers and lakes on day 2, land from sea on the earth of day 3, and Sun from Moon (the coagulation of previously diffuse light) in the heavens of day 4. I then view days 5 and 6 as, in part, emplacements (more additive than differentiative) into appropriate surroundings but also as the final differentiations of each realm--as water, air, and land all bring forth their appropriate living expressions.

Other interpretations abound, of course. In one popular scheme, advocated in two books that I have read and in several letters received from readers of these essays, the six days of creation fall into two equal cycles: three days of preparation followed by three days of population (of the heavens by Sun, Moon, and stars on day 4; of the sea and air by swimming and flying animals on day 5; and of the earth by terrestrial animals and humans on day 6). In this reading, one might view the first three days as differentiative (whereas I would interpret the first four days under this theme) and the last three as primarily additive. Nonetheless, however one explicates the story, I don't see how our usual reading of progressive addition for all six days can possibly be supported. At least the first three days--probably the first four, and perhaps all six--must be reconceived as a creation myth based on the great alternative theme of differentiation from unformed potential, rather than addition piece by piece.

This contrast of differentiation and addition as the two primary modes of organizing stories about sequential and progressive development becomes relevant to students of natural history both as a framework for analyzing our oldest classics of the discipline (the creation myths in our earliest historical documents) and as a guide to understanding our current problems and conflicts. In particular, when we recognize that we derive our concepts of history not only from the factual signals that scientific research has extracted from nature but also from internal limits on the logical and cognitive modes of human thought, then we can appreciate the complex interaction of mind and nature (or inside and outside) that all great theories must embody.

The ancient creation myths of our cultures become particularly interesting in this context, because they originated when our ancestors possessed no direct data at all about the actual pathways of life's history as revealed in the fossil record. These myths therefore represent nearly pure experiments in the range of mental possibilities for explaining the natural :world when no hard information constrains our field of speculation.

The bottom-line or take-home message--that mind and nature always interact to build our basic concepts of cosmic order--becomes especially relevant in our current scientific age, when prevailing beliefs about the sources of knowledge lead us to downplay the role of the mind's organizing potentials and limits, therefore leading us to regard our theories of nature as products of objective observation alone. In particular, the logical restriction of tales about sequential and progressive development to two basic modes--differentiation and addition, in the terms of this essay--can help us to understand our current theories (and to probe their scientific weaknesses when our mental preferences have hidden a different factual reality).

Consider the two major processes in biology--embryology and evolution--that, as applied to human history, must be expressed as tales of sequential development toward greater complexity. (I do not believe that either process, especially evolution, must yield stories in this mode, but I do not challenge the pattern for our own particular case.) Both the history and the current variety of views on human embryology and evolution may be regarded, without gross caricature or oversimplification, as one long exercise in the interplay of shifting preferences for stories about differentiation or addition.

The study of vertebrate embryology, from the invention of the microscope in the early seventeenth century to our modern understanding of genetics, has featured a set of conflicts between differentiative and additive interpretations for a process that features increasing size and complexity, from a tiny, homogeneous egg to a neonate with all the anatomical intricacies of adulthood. From the seventeenth through the early nineteenth centuries, the debate between "epigenesis" and "preformationism" virtually defined the territory of study for embryology. The epigeneticists embraced an additive model, arguing that the initial egg should be interpreted as its literal appearance suggests--that is, merely as a mass of promiscuous potential, devoid of structure and eventually shaped to the particular anatomy of the complex neonate only because formative principles then operate upon this initial homogeneity to build, step by step and in an unerring manner (so long as embryology follows its normal course), the complexity of the final product. ("Epigenesis" means, literally, generated upon--that is, one step after the other.)

By contrast, the preformationists rallied behind a differentiation story that envisaged all the structural complexity of the neonate as already present within the initial cell and only brought to visibility during embryological development. In the caricatured version, preformationism has usually been ridiculed as the belief that a perfect homunculus lies within each sperm or egg cell. No serious scientific preformationist held such a view. Rather, they argued, all structures must be present in the initial cell--but in too tiny, too transparent, and too diffused a state to be visible (like the chaos at the outset of Genesis 1, not like a homunculus). The development of the embryo then becomes a differentiative process of concentration, coagulation, solidification, and growth.

When evolutionary ideas pervaded embryology in the nineteenth century, the two leading interpretations continued to uphold contrasting stories of addition or differentiation. Ernst Haeckel's famous theory of recapitulation held, in a purely additive account, that sequential steps in embryonic complexity repeated the evolutionary accretion of successive adult stages--so that a complex animal, in its embryology, literally climbed its own family tree. The primary alternative, Karl Ernst von Baer's theory of differentiation, argued that the visual simplicity of an early stage does not represent an ancient ancestor that must then be augmented (as in Haeckel's additive theory) but rather a more general form of greater homogeneity and lesser differentiation, holding all potential for the definitive complexity that eventually develops in each lifetime. Thus we know, at an early stage of its development, that the embryo will become a vertebrate, then (at a later stage) a mammal, then a primate, then a hominoid, and finally a human being--a process of increasingly finer specification, contrasted with Haeckel's additive model of ever increasing complexity in accretion.

In the other historical process that led to our human form--the much longer evolutionary construction of Homo sapiens in geological time, rather than the embryological generation of each individual Homo sapiens in nine months--concepts of evolution may also be classified into additive and differentiative models. Darwinism embodies an additive view. Because the Darwinian style of explanation has prevailed within science, we tend to forget that several abandoned theories of evolution advocated differentiative models. These accounts imagined that the first vertebrate, more than 500 million years ago, already contained all the parts and potentials that evolution would necessarily elaborate into human form in a distant future. The supposed mechanisms for such a "programmed" differentiation spanned the full gamut from God's direct actions (in a few overtly theological accounts) to principles embodied in unknown, but entirely physical, laws of nature (for some atheistic versions, at an opposite speculative extreme).

If allegiance to an additive or differentiative model implied no consequences for skewing our views of life in disparate directions, then we could blow off the entire subject as an effete intellectual game. But our preferred theories act as biases that strongly influence our basic conceptions of the natural world. Additive and differentiative views of historical sequences do not hold the same intellectual weights, properties, and implications. We might summarize the differences, looking at lessons from the history of science, by saying that each basic model features a defining property and struggles with a major problem.

Stories of differentiation work primarily from the inside out. That is, the sequence begins with all eventual results already preformed, albeit unexpressed, within an initial homogeneity. How, then, can this potential become actualized? In explicit contrast, stories of addition operate primarily from the outside in. That is, the sequence begins with truly unformed stuff, promiscuous potential that could be drawn into any number of pathways by outside forces. How, then, can such a gloppy mass be carved by external agents into such an exquisitely complex final product (an even worse problem for embryology than for evolution, because the carving must follow the same basic path each time for normal embryos within a species, whereas evolutionary results only arise once)?

In the major distinction between the two models, stories of differentiation work better for determined systems in a predictable world, whereas stories about addition hold the conceptual edge in a contingent world, where any historical sequence may follow innumerable (and unpredictable) options, and the actual result then proceeds from the particular set of external prods that the rolling ball of promiscuous potential (if I may be excused this metaphor) happens to encounter in its trajectory through time. For this primary reason, our modern embryological models tend to be primarily differentiative, and our evolutionary models primarily additive.

After all, embryology does generally follow an internally prescribed route specified not by the preformed parts of preformationists but by the programmed instructions of modern genetic understanding. (We should not accuse the preformationists of stupidity for placing the right idea into the wrong substance. After all, their intellectual world did not include a concept of programmed information, except perhaps as embodied in the old trifle of music boxes or the newfangled invention of the Jacquard loom, whereas no sentient person in our age of genes and computers could fail to assimilate such informational models as an intellectual centerpiece.)

By contrast, the evolution of any lineage wanders along contingent and unpredictable paths of a uniquely complex history. The few lineages, including our own, that do become more complex through time may add their increments of sophistication in a sequence that makes sense after the fact. But even an omniscient observer could never designate, for certain, the next step in an unpredictable future. Therefore, as a description of evolution, additive models that introduce sequential steps from the outside work better than differentiative models that must envision an entire future already implicit and enfolded within any current form.

Under this analysis, we should not be surprised that Genesis 1, despite our usual and unconsidered readings, tells a tale of differentiation rather than addition. After all, if God proceeded with the usual care and thought conventionally attributed to his might, he probably had a pretty accurate idea about the finished product even before he began the work. Biological evolution, on the other hand (at least as viewed under the limits of our eminently fallible mental machinery), seems to wander along a wondrously erratic set of specific pathways within its broad predictabilities.

Our preferred intellectual models do make a difference, and we must therefore be sensitive to the disparate implications of additive and differentiative models as we struggle to understand the history of life. Still, I think that any passionate and curious person can feel the same emotional thrill emanating from either intellectual interpretation. We live in one helluva fascinating universe, whatever its modalities of construction. Thus, if I may beg one last indulgence from my readers (this time for ending with the same image that I invoked just two essays ago, in a different treatment of Genesis and evolution), I happily embrace the common sentiment behind two maximally different views of organic order: the differentiative model of Genesis 1, with its ending of sublime satisfaction: "And God saw every thing that he had made, and, behold, it was very good"--and the additive model of natural selection, so lovingly described by Darwin in the last paragraph of the Origin of Species: "There is grandeur in this view of life."

Stephen Jay Gould teaches biology, geology, and the history of science at Harvard University. He is also Frederick P. Rose Honorary Curator in Invertebrates at the American Museum of Natural History.

COPYRIGHT 2000 American Museum of Natural History
in association with The Gale Group and LookSmart. COPYRIGHT 2000 Gale Group