9 - Four conclusions

Summary

The identities of the gene

The gene today has several identities which have accumulated as the molecular biosciences have developed and diversified. The gene is still an instrumental unit for genetic analysis – the Mendelian allele. The gene is also a material unit of heredity, a reasonably clearly defined structural unit used in annotating genomes – the nominal gene. The gene is also a unit of Crick information, a volume in the ‘library of specificities’ (Nanney 1958). But the relationship between this identity of the gene and its identity as a structural unit has become increasingly vexed in recent years. So the gene as a unit of Crick information has taken on a separate identity – the postgenomic gene. The gene also has less prominent identities: we saw that some ‘genes’ are no more than hypothesised anchors for the parameters of developmental models – ‘abstract developmental genes’. Each of these identities plays a productive role in some forms of biological research. Scientists are adept at thinking about genes in whichever way best suits their work, and at switching between these different representations of the gene as the nature of their work changes. The concept of the gene is therefore best thought of as a set of contextually activated representations.

In Chapter 2 we described how the gene of classical, Mendelian genetics had two identities, as an instrumental unit for genetic analysis and as a hypothetical material unit of heredity. In Chapter 3 we described the elucidation of the basic structure and function of DNA. This represented the successful conclusion of the search for the gene as a material unit of heredity. However, the causal role of the gene as it had been envisaged in classical genetics was very substantially revised in order to fit what had been discovered about the material basis of heredity. Moreover, the original role of the Mendelian gene continues to define the gene whenever biological research uses genetic analysis. As we showed in Chapter 3, it can be necessary to think of genes as both Mendelian alleles and molecular genes, even when those two identities do not converge on the same pieces of DNA.