Structural Constraints, Spandrels, and Exaptation 1221
the development of species with a many faceted spheroid tumbling over from
one facet, or stable equilibrium, to another. The existence of internal
conditions in animals corresponding with such facets is denied by pure
Darwinians.Mivart recognizes that any saltationist claim must resolve the two cardinal
objections that lead many scientists to embrace the gradualism of Darwin and Lyell:
(1) How, short of invoking something miraculous, can so many parts be altered
together, and all at once, to produce a harmonious and adaptive result in a highly
modified descendant; and (2) how can any theory of one-step transformation explain
the parallel or convergent modification of sets of coordinated features, as found in
many independent lineages?
Mivart faces these difficulties along the standard structuralist route, by calling
upon the other concept so often twinned with claims for saltation: internally
channeled change. If alterations, like isotropic Darwinian mutations, could modify an
organism in any direction, then the difficulties stated above would become
insuperable. But if jumps can only occur along certain limited routes, set by the
internal structure of organisms and predisposed towards harmonious alteration of
coordinated parts, then saltations become both limited in directional expression, and
biased towards workability. Again, Galton's Polyhedron suggests such a linkage of
saltation (facet flipping) with internally limited directionality (restriction of routes of
change to positions underlain by adjacent facets). Mivart states (1871, p. 143): "All
these difficulties are avoided if we admit that new forms of animal life of all degrees
of complexity appear from time to time with comparative suddenness, being evolved
according to laws in part depending on surrounding conditions, in part internal—
similar to the way in which crystals (and, perhaps from recent researches, the lowest
forms of life) build themselves up according to the internal laws of their component
substance, and in harmony and correspondence with all environing influences and
conditions."
But what operational good can emanate from an invocation of such internal
forces (even granting the logical soundness of the argument), if the nature of these
forces remains unknown and mysterious, thus reducing their status to special
pleading? Mivart even cites the classical literary spoof of such foolish arguments, as
presented by the phony doctors of Moliere's hypochondriac: "But it may be again
objected that to say that species arise by the help of an innate power possessed by
organisms is no explanation, but is a reproduction of the absurdity, I'opium endormit
parcequ'il a une vertu soporifique (p. 230)" (opium puts you to sleep because it
possesses a soporific virtue).
In reply, Mivart points out that we also know nothing about the physical nature
of Newtonian gravity, but still find the concept useful because such mathematical
regularities as the inverse square law have explanatory, predictive and integrative
power. Similarly, we may not know the actual workings of heredity, but we can, by
empirical cataloguing and experimentation, determine sets of observed regularities in
the variations of modern species. Following the traditions of 19th century
structuralism, Mivart recommends the