Genetic Analysis 283of the genotype-phenotype relationship (not only in the eyes of the lay public).
Only toward the end of the twentieth century, when developmental biology became,
on the one hand more molecular, developmental genetics became also methodolog-
ically “physico-chemical” (see Gilbert [1996; 1998; 2000]), while on the other hand,
with genetic analysis conceptually becoming more system oriented (see e.g., Oyama
et al.[2001]), aided by technologies of handling enormous amounts of data of com-
plexity, the two disciplines merge again. It would, however, be a mistake to claim
that classical genetic analysis neglected the functional and developmental aspects
of organisms: Rather, it concentrated on these aspects from the perspective of
specific genes.
Already in 1902, consulting Bateson, Archibald Garrod, who was interested in
the metabolic aspects of human conditions, recognized alkaptonuria as an inborn
metabolic variant inherited as if it were a recessive Mendelian unit character [Gar-
rod, 1902]. This method of recognizing conditions as “markers” of Mendelian unit
characters allowed Garrod within a short period to identify four human unit char-
acters, and point at them as “inborn errors of metabolism” [Garrod, 1908]. (See
also Harris [1959].)
One of the early issues of gene function was that of the role of dominance.
Whereas Mendel described the dominance of one of the two alternative forms of
each of the seven selected factors as a mere empirical fact [Mendel, 1866/1966,
10-11], de Vries considered it as a major principle property: “The lack of tran-
sitional forms between any two simple antagonistic characters in the hybrid is
perhaps the best proof that such characters are well delimited units” [de Vries,
1900/1966, 110]. De Vries and Hertwig, like Roux and Weismann argued for pre-
formed determinants for all characteristics in the egg, but unlike the latter did
not believe that differentiation occurred by the unequal distribution of these de-
terminants to the blastomers. According to de Vries differentiation was achieved
by activation-inactivation of latent determinants, or pangenes. Correns could not
understand “why de Vries assumes that in all pairs of traits which differentiate
two strains, one member must always dominate” [Correns, 1900/1966, 121], and
suggested instead a physiological, quantitative, rather than the preformational in-
terpretation for the phenomenon [Correns, 1924a, 330]. But it was Bateson who
gave the influential developmental interpretation to dominance in the form of the
Presence-and-Absence hypothesis, recessiveness being absence or loss of thechar-
acter. It must, however, be kept in mind that although he did not distinguish
between character and factor (or phenotype and genotype), yet, conceptually he
considered the development of the organism and he and his coworkers justified
the dealing with a single trait by heuristically assigning all the remaining trait
to the “background”: “[T]he Mendelian contrasting pair, yellow and green might
be regarded as presence and absence of yellowon a basis of green” [Hurst, 1906,
119 my emphasis]. This exactly was also the policy of Morgan, when he realized
that his embryologically-appropriate nomenclature must be neglected for practical
reasons: Methododlogically consider the mutant effect of the gene attended, but
conceptually mind also the impact of the rest as the “residuum”. (For details see