Knowledge, Language, and Coordination
Language is the ultimate archetype for emergent phenomena. Its structure is the archetype for the structure of knowledge embodied in those phenomena. Nobody plans language. Attempts at planning are worthy of the cynicism they receive. Remember the zealous high school teachers and college professors who preached the gospel of MLA formatting? This privileged group of experts may do their best to enforce homogeneous belief upon those who appear to be their subjects, but the emergent features of language remain.
Those who attempt to codify the structure of language are attempting to organize existing data. New data may challenge this structure. The development of language is open-ended as those who use a language constantly generate novelty not captured ex ante by a set of rules. This is true not only for spoken language, but the language of our lives. Just as every word contains meaning, so too does every action. The meaning of and rules governing action often change depending upon context. A sentence spoken or action taken in one arena may receive praise, but in another may draw criticism or even be met with violence. What gives?
For an answer, we can look to Ludwig Wittgenstein. Wittgenstein was a second cousin of Hayek. Hayek reflected upon his experience with Wittgenstein:
One had to ‘live’ truth and not tolerate any pretence in oneself or others. It sometimes produced outright rudeness and certainly, unpleasantness. Every convention was dissected and every conventional form exposed as fraud. Wittgenstein merely carried this further in applying it to himself.
Wittgenstein sought a general structure to interpret social reality. He came to describe human interaction as comprising language games (1953). We can think of language games as call and response, much like when two jazz musicians solo back and forth with one another (those not acquainted with jazz may think of a “guitar battle”). One phrase may invoke a multitude of phrases in return from the receiving actor. Some phrases may work better than others. Some phrases may sound better to some audience members than to others depending on their preferences for and familiarity with certain types of music. The interaction between the musicians will themselves be shaped by their own knowledge of and preference for music as well as past experiences with one another, both musical and non-musical. The play taking place – itself a language game – is embedded in an array of language games.
Just as notes comprise phrases in music, so too do actions comprise phrases in language games. All words and actions convey meaning that are derived from their context. This holds on multiple levels. Change one word in a paragraph and you may instigate a multitude of changes in meaning. The same holds for the meaning imbued upon any action. From the view of the actor, one meaning may result from his or her available set of knowledge, while a totally different meaning may be imbued by an observer who carries a different set of knowledge, even if that knowledge is only marginally different.
The social scientist and the historian face this hermeneutical problem at all levels. All agents interpret their environment. This includes interpretation of actions of other agents and these actions’ effects. Knowledge is not given. Meaning is not obvious. Consider the texture of knowledge that exists within different firms in the same industry. While each firm may approach the same type of objects and processes, the culture of any two firms may be substantially different. Members of each firm likely follow unique sets of rules. The words and phrases used to communicate, commitment to and timing of social gatherings, and expectations concerning the timing of the completion of certain tasks may all be different (Grant 1996; Cremer 1993; Arrow 1974). Human psychology promotes such homogeneity so this feature is inevitable (Asch 1955; Sowa 2007; Burt 1999; Henrich and Boyd 1998). If workers and management judge the actions of a new comer according to the system of meanings that they have already developed collectively, it is possible that they and the uninitiated employee may experience confusion in light of asymmetric interpretations (Klein 2012). Successful entrepreneurs within firms will overcome dissonance and uncertainty created by apparently conflicting mental models and thus promote collaboration even when disagreement exists (Koppl and Langlois 2001; Kirzner 1997; Knight 1921).
Language games represent the dynamic links between the agents that comprise a group. Those who are able to mutually participate in these language games exhibit knowledge that is shared or in some way compatible. The boundary of a group is coterminous with the boundary of some collection of games. We define a social group as an entity comprised of agents who collectively experience “a network of meanings” (Hassard 1994, 307; Mauws and Phillips 1995) in regard to shared language and “a network of significations” in regard to relationships that comprise some organizational structure (320; see also Danford 1978, 73-121). The concept of a mental model, popularized amongst economists by Denzau and North (1994; Hayek 1943), carries with it substantial content that is necessary for the functioning and description of any social world.
While some games are confined within a given network, others reach across networks. The language game that contains perhaps the broadest reach is the game of market exchange and pricing. The first person who appears to have recognized this relationship is Saul Kripke (1982, 112-113, n89). Others include David Bloor (1997) and Roger Koppl (2002). Kripke notices that there seems to be “a certain analogy between Wittgenstein’s private language argument and Ludwig von Mises’s celebrated argument concerning economic calculation under socialism.” Prices are set, not by a central node, but by conversation between agents who participate in the market process (Mises 1920; Hayek 1935; 1945). Mises and Hayek recognize this fact. David Bloor elaborates:
Thus: ‘Prices are . . . social phenomena as they are brought about by the interplay of the valuations of all individuals participating in the operation of the market’ (HA: 331). Price is not like, preference, an individualistic fact, but a collective fact derived from specific form of collective organization. (1997, 75)
Prices are transformed by the interaction of market participants. Agents engage in dyadic interactions that reflect their beliefs and preferences (Wagner 2010; 2016). When bargaining agents come to an agreement, they have converged on agreeable values of the goods being exchanged. Just as groups tend to converge on sets of common knowledge, participants in the game of market pricing and exchange converge upon a bounded range of prices. This game is played by agents from just about every group imaginable. To the extent that mutual participation in a language game represents a “social bond”, the game of market pricing is certainly the most inclusive game, connecting agents and groups both near and far.