Knowledge must be very differently organized in an oral culture than it is in one with writing and, of course, memory is the key. People remember through time, and the memory of an individual is limited in extent. A society may organize itself in such a way as to maximize the common store of what is remembered and may also find ways of setting aside those matters that lie outside its memory range. In this article I aim to formulate a descriptive model for a society that operates in terms of what I call a "memory capsule" of four generations that provides an expectation of recollection over a period of about a hundred years. This model represents a rather static way to speak of the actual human experience that moves forward constantly through time as each generation replaces the one before it, and we should take account of this process, but I have found through the study of Indo-European material, and comparable material from elsewhere, that it is possible to envisage this model in terms of movement through a system of alternate generations, with each of the alternations having its distinctive nature. In the Indo-European case, it is proposed that the generations are marked by the central institution of kingship, with each king's reign differing in nature from that of his predecessor and successor. The "stepping stones" of my title suggest a way of grasping this proposed dual movement through time.
Focus on the Oral-Cultural World
The stress on the written word in the West has meant that we are only belatedly looking into the nature of the oral-cultural world that must have existed before writing came along. It could be argued that scholarship has long taken the historical approach to the Indo-European past too much for granted, and moving toward an oral approach calls for a radical and much more rewarding shift in perspective. A historically known society does not just come out of nothing but has a prehistoric past. We cannot trace this past in any detail before the advent of writing, but what we can reasonably do is build a spatiotemporal model of the posited prehistoric cosmology, resting on folk material and on scraps of information from earlier times (Lyle 1990, 2006, 2007, 2010). It is an orally organized cosmology that provides the foundation from which the diachronic developments that we can document took their rise. The oral-cultural elements found in Indo-European societies are not simply offshoots of those processes for which we have early written evidence, but belong to a free-standing oral base to which written elements were later attached. It seems well worthwhile to put effort into grasping the nature of this base. As Walter Ong commented (1982:13): "You cannot without serious and disabling distortion describe a primary phenomenon by starting with a subsequent secondary phenomenon and paring away the differences." We have to attempt to explicate the primary phenomenon of a posited cosmological society in its own terms.
In undertaking study of this kind, it is helpful to consider the operation of oral, or largely oral, societies throughout the world and to use comparative methods of research, as well as seek clues within the wide Indo-European culture area. It should be noted that, even when it is Indo-European evidence that gives the possibility of historical depth, the results may be found applicable to other cultures as well.
The current essay pertains to a specific scenario with regard to time, and its basic idea is that people without written records may apprehend a limited segment of linear time and may work to organize and control it in an optimal way. They can achieve such organization by using the time measure immediately available to them, that of human generations. However, because of the wide spread that is possible between births from one couple, a socially agreed upon means of determining the length of a generation is required. In the Indo-European model, a generation coincides with a king's reign that is constrained by an age-grade system and can be postulated as lasting twenty-four years (Lyle 1997). In...