"For the tool-combination principle is indeed a law of progress."
Clarence E. Ayres (1944, 119)
The Problem: Are there Universal Principles of Knowledge Evolution?
In the first half of the 20th century, institutional economists had proposed a specific theory about technological progress that differs considerably from currently held convictions in economics. Veblen and Ayres saw a fundamental tension between the knowledge that is evolving in technology and the rules that govern human social interaction (for a more recent view, see Bush, 1987). There are two theoretical cornerstones of this theory. The first one, very much emphasized by Ayres (1944, 125ff.), is the assumption that technological evolution proceeds almost autonomously from individual human action. The second one is the analysis of social competition among human beings, with the latter being seen as hampering the former.
In this brief paper, I show that these two ideas can be put into a more abstract and general framework that draws on recent advances in the life sciences and philosophy. The starting points of the argument are:
* The evolution of knowledge can be understood independently from conceptualizing knowledge as mental content.
* Knowledge is a system of rules, which have physical correlates.
* The evolution of these rules follows principles that are cross-domain in the sense of encompassing natural/physical and cultural/mental phenomena.
The principles find expression in specific ontological constraints that hold for the concepts of knowledge and information. This approach is close to "Universal Darwinism" as propagated by Hodgson (2002) in economics (for an elaboration on the differences, see Herrmann-Pillath 2007b).
De-humanizing Knowledge in Economics
In contemporary economics, there is no unified use of the concepts of information and knowledge. In economic theory, they are mostly related to the role of information in the coordination of individual actions, such as "common knowledge" in game theory and "asymmetric information" in the mechanism design literature. Such an approach produces a heavy bias toward intersubjective reflective knowledge, i.e., ego's knowledge about what alter knows about ego's knowledge, and it concentrates on mental content of individual human beings (Samuelson 2004). It is almost totally disconnected from the uses of the concept of knowledge in the theory of economic growth, where it is conceived as a stock of productive knowledge (e.g., Jones 2002). Few uses point toward an integration, as in North's (2005) approach, where knowledge processing in the economy is seen as being increasingly focused on the coordination of transactions. North posits that institutional evolution supports technological progress.
Knowledge certainly plays a central role in recent vintage evolutionary economics, mostly focusing on the mechanisms that produce new knowledge, which entails that the individual epistemic agent is moving out of the center of analysis because of the transfer of Darwinian concepts (Hodgson and Knudsen 2006). This elaborates analytically on earlier uses of terms such as "tacit knowledge" where a firm is conceived as a set of rules and routines, which carry knowledge that is not necessarily accessible to the individual members of the firm (Nelson and Winter 1982, 96ff.). Consequently, the evolution of the firm can be conceived as resulting from the selection of routines (Knudsen 2002). Similar views can be found in recent evolutionary approaches to technology, which also treat knowledge partly independent from the human agent (e.g., Mokyr 2000). Thus, without making this relation explicit, contemporary evolutionary economics revives the idea of dehumanizing knowledge that was implied in the views on the "machine process" and technological progress held by the early institutionalists. The emerging theoretical framework is evolutionary theory in the Darwinian sense, which interprets knowledge as an evolving population of rules.
Turning Information into an Ontological Category
This trend of de-humanizing knowledge in evolutionary economics meets the trend to universalize the concept of information in the sciences. The "information-theoretic turn" has now reached even physics (Baeyer 2003). Information is a core concept in biology (Maynard Smith 2000). This observation raises the question whether and how some universal features of information can be defined and specified that impose constraints on the uses of these concepts across all scientific disciplines. I call these universal constraints "ontological," thus accepting the view that is emerging in physics that information is a fundamental feature of the world (Brukner and Zeilinger 2005), and that "information is physical" in the sense that there is no information without physical substance (Aunger 2002, 136ff.). My conception of ontology follows Bunge (1977), that is, ontology is not treated as an irreducible axiomatic system, but as an abstraction from empirically grounded knowledge about basic structures of the world.
Subsequently, I refer to both information and knowledge, intuitively treating the first as a flow and the latter as a stock. Information is what is added to the stock of existing knowledge, a use that is familiar in economics, which however, is not prevalent in the other sciences that concentrate on information. Knowledge is commonly related with the human mind in the sense of a particular state of mind, which is described as "knowing that." By contrast, information can be independent from the human mind, as for example, the use of "genetic information" illustrates (Adams 2003; Floridi 2004). However, precisely this conceptual asymmetry shows that there is a need for conceptual...