Novelty in Evolution: Introduction

The scenario is conventional: a mutation occurs, which results in a new_0 gene sequence coding for a new_1 protein whose interaction with the chemical machinery of the cell, set up by the remaining gene products, triggers a cascade of new_2 chemical reactions resulting in a new_3 signaling circuitry which enables the utilization of new_4 information.

Each time the word ``new'' denotes a different kind of novelty, because each time different kinds of constraints are in effect:

0. 

novelty_0: A sequence is a simple combinatorial object. At this level the generation of novelty is virtually unconstrained. If the sequence is sufficiently long, any random replacement of any symbol at any position yields a new_0 sequence. Novelty_0 is a throw of the dice.

1. 

novelty_1: A protein is more than a sequence of symbols. It is a sequence that folds into a shape as a consequence of interactions between symbols along the chain. Three-dimensional space and the nature of intramolecular forces constrain which shapes are possible. At chemically relevant levels of resolution these constraints result in considerably fewer stable shapes than sequences. Not every novelty_0 is a novelty_1.

2. 

novelty_2: The types of chemical groups and their position within a molecule define its ``domain of interaction'', that is its capacity to participate in specific chemical action. Novelty_2 is a matter of chemistry.

3. 

novelty_3: The constraints and opportunities of interaction within a given network of chemical pathways determine which new_3 network roles a new_2 molecular agent can participate in. How (or whether) a network forms or changes, depends on its molecular components, their interactions and kinetics. Novelty_3 is a network property.

4. 

novelty_4: The innovated_3 network is characterized by its molecular components and their relationships. What is regarded, however, as new_4 ``information'' is a matter of the coupling between this network and other such networks either within the same, or between it and other, levels of biological organization. Indeed, it is the joint construction and maintenance of a chemical reality composed of a large number of linked networks which defines the biotic element of an environment. Novelty_4 is a property of a network of networks.

It is plain that novelty_4 cannot occur unless novelty_0 occurs. There is, however, a gap between novelty_0 and novelty_4 that theory is presently unable to bridge. We understand the possible for novelty_0 as the space of sequences over an alphabet. Yet we are unable to even specify the nature of the possible for novelty_4. The problem is to understand the relation between genotype, the trivial level of novelty, and phenotype, the suite of organizational levels and associated behaviors generated by interactions among gene products. This relation is referred to as development. Here I use the word development to denote any processes that connect ``vertically'' two or more layers of biological organization.

Evolution is roughly the conjunction of two factors: the selective amplification of genotypes based on the differential reproductive success conveyed by their phenotypes, and the modification of phenotypes through chance events at the level of genotypes. The theoretical agenda pertinent to the first factor aims at characterizing the conditions under which an innovation can, once generated, invade an existing population. The theoretical agenda pertinent to the second factor consists in understanding how novelty is generated in the first place - in other words, to understand the possible in biology by characterizing the routes along which one phenotype can be transformed into another. Virtually no theoretical framework exists to address the second agenda. Yet, without such a framework we cannot make sense of evolution, because we cannot understand its outcomes. To paraphrase Leo Buss, the hard task of evolutionary theory today is not to recognize an innovation as advantageous in some fitness measure, but to understand how an innovation generates opportunities and constraints for subsequent innovations. This is not idle theorizing, it is use-inspired basic research: Without such a framework, we cannot hope to achieve anything that deserves the name of biotechnology, if technology means reducing the degree of empiricism in a practical art.

The overall theme is to explore notions of organization at vastly different levels of abstraction and to understand innovation for each organizational class in response to change ocurring at the level upon which that class is founded. The realm of molecular organization is the one where substantial theoretical progress is most likely to occur in this decade. Yet, the evolution (in some generalized sense) of functional, self-maintaining and homeostatic organizations is a central theme in many fields beyond biology proper, including many cognitive processes, as well as a diversity of human social and economic institutions. The origination and innovation of organization constitutes a ``vertical question'' cross-cutting chemistry, molecular biology, cognitive science, social science and economics. In all these disciplines the challenge is to (1) achieve a clear definition of organization as an autonomous individualized entity distinguished from mere aggregation, (2) understand how the robustness required for autonomy and individualization squares with evolvability, the capacity to be innovated, and plasticity, the capacity to be flexible and adaptive without losing identity, (3) understand the topology of the possible, that is, the routes by which organizations are transformed into new organizations, (4) understand what determines the possible.

I briefly introduce and review some items from my own research in an attempt to illustrate ways in which the scientific agenda outlined above can be approached.

Walter Fontana, Santa Fe Institute