A Cognitive Archaeological Framework

When did Cognitive Archaeology become a field of archaeology in its own right? What exactly does it cover and what relation has it to the study of art and especially symbolism?

"Cognitive science is an interdisciplinary approach to studying mind and, in particular, intelligent thought and behaviour. It is a relatively new academic discipline which is developing from a merger of interests among certain linguists, psychologists, philosophers, computer scientists, anthropologists, neuroscientists and others" (Norman 1981, as cited by Segal 1994:22).

The mainstay of this dissertation is a study of the symbolism created by Palaeolithic Man, and the effect the methods of its generation had on the future creative processes of producing art. What did art originally mean, if anything, and how do the reasons for producing art differ from more recent motives? This is therefore linked to the thought processes of our forefathers. The most recent discussion of such ancient cognitive processes has been dealt with in a book entitled "The Ancient Mind" (Renfrew, C. and Zubrow [eds.] 1994). Renfrew (1994:xiii) notes that, as processual archaeology revolutionised archaeology in the 1960's and 70's cognitive archaeology will revolutionise the discipline in the 1990's, and even part of the 21st Century. According to Renfrew (1994:xiii) cognitive science is in its childhood and cognitive archaeology is in its infancy. Renfrew et al (1994:xiii) believe that a Scientific approach, encompassing scientific tradition and empirical methodology (Cognitive, mathematical and computer sciences) as opposed to a Hermeneutic, semiotic or a Interpretationist, anti-scientific literary approach begins to lay down the foundations for a science of cognitive archaeology. In "The Ancient Mind", five main issues have been addressed:

what are the trends in artificial intelligence, cognitive psychology and cognitive anthropology that are applicable to cognitive archaeology?
what is the present level of theory in cognitive archaeology?
what tools and scientific methodology are necessary for cognitive archaeology?
what problems are amenable to solutions given the present state of cognitive archaeology? and
how do different scholars in different cognitive fields examine archaeological data in order to make cognitive inferences?

Renfrew (1994:5) points out that the intention of the term "The Ancient Mind" is not to imply that there is something inherently different between the thought processes of yesterday and those of today. No distinction is implied between the ancient mind and the modern mind. Nor is it implied that there were a series of evolutionary stages in human cognition. Renfrew et al (1994:5) make no assumptions about different kinds of categories of thought (there is no way of knowing whether such differences existed and therefore no assumptions can be made about them).

However, Renfrew (1994:5) notes that we cannot overlook the fact that there may be differences between the thought processes employed by modern humans and those of our ancestors recent and distant.

The scope of cognitive archaeology could be outlined in several ways (Renfrew and Bahn 1991: 339-70, as cited by Renfrew, 1994:5). Perhaps the most concise approach is to focus explicitly upon the specially human ability to construct and use symbols. A symbol is something which stands for or represents something else: "a visible sign of an idea or quality or of another object" (Webster's Collegiate Dictionary 1925:974 'symbol'). The word derives from the Greek 'to place together' and the notion of juxtaposition (of x against y), of representation (of x by y) and of metaphor (where x is equated with y) are closely related.

"An important component in the cognitive-processual approach is to set out to examine the ways in which symbols were used. This may be contrasted with an attempt to seek rather to ascertain their 'meaning', which would generally be the object of the anti-processual or interpretive approach. The distinction is an important one. As we shall see, both approaches must inevitably rely upon the insights and intuitions of the modern investigator. The creative, and in that sense perhaps subjective aspects of scientific inquiry are not in doubt, and it is a common misconception of the scientific method deriving largely from the polemic of Bourdieu, that it is, in its inspiration, inhuman, mechanistic or lacking in creativity. But for the cognitive-processual archaeologist, it is enough to gain insights into how the minds of the ancient communities in question worked and into the manner in which that working shaped their actions. For the interpretive archaeologist, working in the grand tradition of idealists like Collingwood, it is not enough. One seeks instead to 'enter the mind' of the early individuals involved through some effort of active empathy. This total experience of 'being' that other, long-dead person, or at least undergoing an experience to be compared with theirs, is what characterises the subjective, idealist and interpretationist approach of the anti-processual and 'post modern' archaeologist. The cognitive-processual archaeologist is skeptical too of the privileged status which must inevitably be claimed by the idealist who is advancing an interpretation on the basis of this intuitive 'I-was-there' experience" (Renfrew 1994:6).

Bell (1994:15), corroborates this by saying that archaeologists should aim at constructing testable theories of prehistoric cognition. Theories about prehistoric thinking should not simply be interpretations. This brings to the fore the interpretation versus testability debate. Testable theories (unlike interpretations) cannot be modified in a way that simply explains away uncomfortable data. Bell (1994:17) goes on to comment on universal statements. Generalised statements are perfectly appropriate and even desirable so long as they entail testable statements. As a matter of fact, generalisation can increase testability because a more sweeping claim can entail a greater number of testable implications. By adopting such a broad area for study and systematically working through the variables, it is possible to refute or confirm their possible role within the whole scheme.

We can now outline the variables and decide how to test them, using questions concerning:

the existence of phosphenes, form constants, and hallucinations;
the evidence of phosphene-like symbols in prehistoric art;
the involvement of drug-induced trance states for the experiencing of hallucinations and visions; and
the availability of drugs in widely varied prehistoric and modern environments.

By way of substantiating these statements I shall endeavour to:

use all available physiological, neurophysiological, and psychological evidence and studies to prove the existence of phosphenes, form constants and hallucinations;
use all available archaeological and anthropological data to build up a body of evidence to test the existence of phosphene -like symbols in prehistoric art;
use all available anthropological data to test the theory of involvement in drug-induced trance states for the experiencing of hallucination and visions; and
use all the available pharmacological and ecological data to test the theory of the availability of hallucinogenic drugs in different environments throughout time.

"I have long maintained that it is not possible to learn about what was in the minds of prehistoric people - these minds are gone, and their mental contents are not recoverable. While we can examine many aspects of what they did, we can understand nothing about what they thought about what they did. But is this really true?.....Although I continue to maintain that we cannot actually know what prehistoric people thought, I now think that it is sometimes possible to make plausible inferences about what they must almost certainly have thought, given very strong circumstantial and analogical evidence. Moreover, I think that some such inferences are testable" (Hill 1994:83).

Hill (1994:85) points out that there are many things which we can never know with any degree of certainty, and that one rarely finds proof of anything in science, outside formal logic and mathematics, and for this reason, certainty should not be expected. Hill (1994:83) goes on to outline two methods for testing data which does not readily succumb to formal, scientific methods of testing. His chosen two methods are summarised here:

i. "Established Generalisation Testing Method" (EGT)
This method involves testing the applicability of an established generalisation against the structure of the archaeological record. Testing set hypotheses will be made more possible where ethnographic links suggest (Hill 1994:83). This brings us to the second method. Hill (1994:83) applies this method to data supplied by Winters (1968) in which he "used data from the Late Archaic in the Midwest to infer that conch shells and copper were considered to be materials of value by the prehistoric peoples involved" (Hill 1994:83).

ii. "Tight Local Analogy Method" (TLA)
This method requires that the unknown (archaeological data) and the known (ethnographic analogy) phenomena be 'tight' and 'local'. 'Tight' meaning that there should be as many specific similarities as possible, and 'local' meaning that they should be close in time and space. However, two problems arise. First, as Binford (1976, 1967) pointed out, we cannot logically study the past by studying the present; we must use prehistoric date if we want to understand what went on in prehistory. Secondly, using analogy in this manner precludes the study of cultural change, since it assumes that there has been no change from the prehistoric to historic context.

Hill (1994:87) applies TLA to data collected by von Gernet and Timmins (1987) on the ritual use of tobacco via parakeet pipes. This data will be discussed in chapter IV, p.64. C It is the TLA model which would be the most applicable to the form of data studied by Lewis-Williams and Dowson (1988). In order for their theory to be accepted, as far as the data will allow, one must endeavour to find a society which engages in:

the taking of hallucinogenic drugs;
the taking of such drugs as part of a shamanistic ritual;
the use of shamanistic rituals in order to enter trance states;
the production of visions during trance which include entoptic forms; and
the relaying of these images, once the trance is over, onto cave walls and pieces of mobile art.

To meet the criteria laid down by the TLA model, the fit must be both 'tight' and 'local'. However, in cases where there is little actual physical data, one could simply alter the theory about the unknown (archaeological data) to comply with the known (ethnographic analogy). Ideally, in chapter V, we shall be looking to uncover a society which complies with the five criteria outlined above, in order to substantiate the theory put forward by Lewis-Williams and Dowson (1988).


On the basis of the criteria laid down by Renfrew (1994:xiii), the theory proposed by Lewis-Williams and Dowson (1988) falls within the boundaries of cognitive archaeology. It uses the research carried out by other disciplines, namely neurologists, psychologists, anthropologists and physiologists. Lewis-Williams and Dowson's (1988) theory is not so much a study of the thought processes of prehistoric peoples as an assessment of the alteration of their thought processes by means of hallucinogenic drugs. Theories concerning their thought processes are still suppositions, but the basis for these theories can only become more stable with more research, and the proposition of possible thought processes should not be discarded because it cannot be proved.

Copyright 1995, Suzanne Carr. All Rights Reserved.


Introduction | Chapter 1 | Chapter 2 | Chapter 3 | Chapter 4 | Chapter 5 | Chapter 6 | Chapter 7 | Chapter 8 | Appendix | Bibliography