Michael E Smith asks why some New Archaeologists adopted Hempel’s covering law model of explanation in the early 1970s despite its known problems. A good question, and definitely worth pondering.
There haven’t been volumes written about this, but there has been some reflection, and clues are available here and there. The short answer is that they adopted Hempel’s model because they wanted to. It was attractive to them because it solved some of the problems they thought they were facing.
Why was Hempel’s model attractive to New Archaeologists?
Hempel’s model was attractive to a particular group of New Archaeologists not because it was good theory, but because it fit their project, and it fit their historical context. Merrilee Salmon (1992:227) drily notes that “The new idea in new archaeology, circa 1960, was to secure scientific status for the discipline”. In a post-war world of rapidly expanding universities and even more rapidly growing science funding (Jaffe 1996: 12659), the New Archaeology faction felt it was important to align with natural science rather than with the humanities. They needed a philosophy of science that would help them do that.
Trigger (1989: 299) recalls that “They were anxious to demonstrate that ethnologists were wrong when they smugly proclaimed that archaeology was ‘doomed always to be the lesser part of anthropology (Hoebel 1949: 436)’. On the contrary, by embracing science, they would surpass cultural anthropology and emerge, as a dominant discipline and a force for social good, from the natural history twilight to which traditional archaeology had been consigned, and which they saw as little better than antiquarianism.
The project and the context
In their particular context of American growth, hegemony, optimism, and belief in progress, the young New Archaeologists wanted their discipline to matter, and they wanted it to matter as much as the more prestigious sciences. Most of all, they wanted to contribute, a theme that permeates Fritz and Plog’s (1970) Hempelian manifesto. “Thus, present archaeological theory contributes less to our understanding of specific human phenomena than many of us would like” (Fritz and Plog 1970: 409). Later, “…we all want our discipline to contribute to the knowledge of laws of human behaviour” (Fritz and Plog 1970: 411). Hempel’s model was to way to get there, because it is “at worst, an important heuristic device which provides insight into the structure of archaeological knowledge. At best it points the way archaeologists must travel if they are to contribute to the corpus of laws of human behavior” (Fritz and Plog 1970: 406).
They were actively and explicitly looking for ways to increase archaeology’s alignment with science, to increase its prestige, and to give it the tools it needed to make real contributions beyond documentation of the past. Fritz and Plog’s choice of Hempel’s model as a tool was motivated by those concerns, which had by then been consistently expressed by a small group of reformers for at least two decades. The opening paragraph of Caldwell’s (1959: 306) The New American Archaeology is worth quoting at length in this regard:
“It is supposed that behind the infinite variability of cultural facts and behind the infinite and largely unknown detail of historical situations we shall discover the workings of a finite number of general cultural processes. This hypothesis underlies much of recent archaeological thought despite the view, often propounded, that because of level, cultural facts are much more complex than those of the physical sciences. This latter assertion does not make our task impossible. Not all cultural facts are of equal importance in determining a given pattern or trend. Certain developmental patterns must surely be overriding in their effects upon other patterns. A major historical pattern may serve to unite or in some cases subordinate other patterns of more limited range”.
Caldwell’s thinking is at the very cusp of systems thinking. It is historically poetic that it is followed in the journal by an article on cloud physics, but it is significant that the word system appears in the article both in its modern and its previously conventional senses . We find it as part of “mid-western taxonomic system”, and also in its emerging sense, as in “hunting-gathering systems”. But the word “pattern”, striving to transcend its culture and personality incarnation, mostly stands in as our modern “system”.
The processes proper to particular patterns must be tested, for example: “…this does not mean that archaeological formulations at the pattern level cannot be tested and that some kind of validation cannot be secured” (:306). This early concern for validation foreshadows Hempel’s focus on confirmation, and may have predisposed the New Archaeologists to see Hempel’s model as the tool they needed, as opposed to available falsificationist philosophies.
Archaeology as science
The project is clear. Archaeology is the science of prehistoric systems. It must proceed by hypothesis, testing, and confirmation. The Hempel-Oppenheim deductive-nomological model, D-N, as it was abbreviated at the time, despite all its flaws, explained to archaeologists in sometimes pedantic detail by philosopher Michael E. Levin (1973), is nonetheless a good heuristic for studying certain kinds of systems (Fritz and Plog 1970:406). One can posit laws that regulate the behaviour of a system, and one can test whether the observed behaviour of the system is consistent with those laws. One can confirm, in the unfortunate parlance of the D-N model, that one understands the system under study. The job was difficult, granted, but not impossible, and, for the sake of the discipline and of society, must be attempted.
The D-N model speaks the language of science. It is to the New Archaeology as the screwdriver is to the paint can. Not designed for it, but close at hand, and gets the job done. For the New Archaeologists, it served one purpose: to open the can. And open it they did.
Caldwell JR 1959. The new American archaeology, Science 129:303-307.
Fritz JM and FT Plog 1970. The nature of archaeological explanation, American Antiquity 35: 405-412.
Hoebel EA 1949. Man in the primitive world, McGraw-Hill.
Jaffe AB 1996. Trends and patterns in research and development expenditures in the United States, Proceedings of the National Academy of Sciences USA 93: 12658-12663.
Levin ME 1973. On explanation in archaeology: a rebuttal to Fritz and Plog, American Antiquity 38: 387-395.
Salmon MH 1992. Philosophical models for postprocessual archaeology, in L Embree (ed), Metaarchaeology: 227-241.
Trigger BG 1989. A history of archaeological thought, Cambridge University Press.