Last Update: 13 April 2004
• • •

Position Papers

The papers collected here are working documents which have not been edited for publication. They should not be quoted without the permission of the authors.

To communicate with the authors via email click on their names.

Ten Positions On Symbolicity In Archeology
Joao Zilhao (Department of Archaeology, University of Koeln)

The origins of symbolling
Robert G. Bednarik (International Federation of Rock Art Organisations)
Picture Gallery

Criteria of symbolicity. Intrinsic and extrinsic formal properties of artifacts
Paul Bouissac (University of Toronto, Victoria College)

The Status of Ethics in Contemporary Epistemology and Ontology, and the Problem of Meanings and Values (the Symbolic) in Archaeology
Stephanie Koerner (School of Art History and Archaeology, University of Manchester)

Stone tool "style" and the evolutionary origins of symbolism
Philip G. Chase (University of Pennsylvania, Museum of Archaeology and Anthropology)

Archaeological data on symbolic thinking in the European neolithic
Eszter Bánffy (Archaeological Institute of the HAS, Budapest)

Stone Age symbolic behaviours: questions and prospects
Andrea Vianello (Graduate School of Archaeology, University of Sheffield)

The Everyday Life and the Symbolism in the Prehistoric Balkans
Lolita Nikolova (University of Utah and International Institute of Archaeology)

Clever Etchings:
Prehistoric language, religious language, and prehistoric religions

Peter Jackson (University of Chicago)

V. Gordon Childe among the “vulgar cognitivists”.
Michael Chazan(University of Toronto, Department of Anthropology)

Symbol for them / symbol for us?
Lisbeth Bredholt Christensen & David A. Warburton (Aarhus Universitet, Denmark)

Printable version of the position papers which were presented and discussed in St. Petersburg at the EAA round table of September 13, 2003

Position Papers (PDF: 390K)



The Problem with Symbolism: Symbols for them, Symbols for us?

Lisbeth Bredholt Christensen & David A. Warburton (Aarhus Universitet, Denmark)


There are several conceptions of the word "symbol". From the basic sense of the word-that of "a sign representing something else"-interpretations vary. Symbols are thus today distinguished from other signs by not being translatable into one, unambiguous meaning. Paul Ricoeur talks about symbols as being elements with "surplus of meaning" (Ricoeur 1976). While this is true today, it may not always have been true. We simplify by suggesting that symbols resist reduction in the way that icons are most effective because reducible. Somehow established by convention, symbols are then imbued with diffuse and unclear associations and meaning.


Using something (e.g., a Cross) as a symbol means that the item in question has its own existence (as a physical object intended for execution), and that its role is not served by its own existence (a utilitarian implement), but rather by the fact that some form of code is shared among those familiar with the item or behaviour in question and its true role (i.e., as representative of a community). This implies a shared ideological understanding.

Symbols are thus social: something can be a symbol only if at least two people agree so.

Although symbols may be both verbal and material one must ask whether a symbol can be a symbol without it having been agreed upon in language. Does a symbol require a mutual verbal agreement between at least two people for something (e.g., a tool) to be also a symbol? It is peculiarly characteristic of scholarly literature that symbols are frequently taken to be basically verbal and dependent upon language, e.g., Lévi-Strauss refers to the "effectiveness of symbols" in terms of words and metaphors without any reference to material objects (Lévi-Strauss 1972).

It is commonly said that symbols are human. It is also commonly assumed that symbols "make use" of language and that language is what separates humans from animals. Is it, however, possible that language is not as old as mankind, but that it did not emerge until the Neolithic and that symbols therefore are no older than that?

In this paper we want to discuss some different issues. Emic-etic aspects of symbolism. Is it possible, or even relevant, to try to get to their conception of symbol? The historical character of "symbolism"; in prehistory we may be dealing with the origins of symbolism. Where did symbolism begin? The language-based character of symbolism. What role does language have? The relevance of distinguishing "practical" from "symbolic" in prehistoric material. Is this possible and/or legitimate?

What is the significance of the explosion of symbols from the Neolithic and the Bronze Age? Can we extrapolate backwards or should we draw a line, putting the Palaeolithic and Neolithic into separate categories when discussing "symbolism"?


The purpose of the present round table is presented as concerning, i.a., "the issue of whether their [early objects] functions were practical or symbolic at the time when they were created" (our emphasis), as well as attempting at answering the question about "Which observable properties of an artifact can count as reliable indications that it once was endowed with symbolic relevance?" (our emphasis).

As such, the aim of the round table is somehow to find a methodologically sound way of getting at "their" - the prehistoric peoples' - concept of symbolism and to be able to reconstruct "their" way of distinguishing between "practical" and "symbolic". Such an aim is in accordance with mainstream anthropology whose purpose, according to the Encyclopaedia of Cultural Anthropology, is ultimately to reach and reconstruct an emic perspective (ECA 1996, Emic/etic perspective). "Emic" is opposed to "etic" and may be defined as a perspective that "focuses on the intrinsic cultural distinctions that are meaningful to the members of a given society" (ECA, Emic/etic perspective, 382). Opposed to this, an etic perspective "relies upon the extrinsic concepts and categories that have meaning for scientific observers" (ibid.).

The distinction emic/etic thus focuses on two perspectives from which to view material: (a) an internal and (b) an external perspective. The distinction emic/etic also puts the finger on the difference between Anthropology and Prehistory, namely that the emic perspective is accessible only via verbal information, and thus inaccessible in terms of the limits of prehistoric material.

Posing questions about prehistoric symbolism and prehistoric distinctions between practical and symbolic is relevant and legitimate. Such queries can, however, only be answered on an etic, and not on an emic level.

"Symbol" is our term: prehistoric peoples did not use it. Talking about "symbolism" in prehistoric material could mean imposing "our" categories onto "them" rather than "us" recognizing and reproducing "their" categories.

These are, however, also the conditions for scholarly work: In attempting to understand the people under study we create categories by which we distinguish phenomena of their culture.

On this basis (prehistoric "symbols" being "our" interpretation of "their" world as "we" see it), we can ask the question: Does it make any interpretative sense to distinguish between practical and symbolic? One can perhaps enquire about the value of distinguishing the "practical" from the "symbolic", any more than the "religious" from the "mundane". These distinctions did probably not exist for the people and the cultures under study.

By contrast, we should appreciate the fact of the challenge that it is Prehistory which offers information about the "origins of symbolism". We cannot deduce from known forms (e.g., crosses) how any given form came into being

We will opt for a more holistic approach, taking material culture as such and as a whole as symbolic and as open to interpretation, without attempting to reconstruct emic categories. This approach is in line with Foucault and Collingwood. Although we need not adopt the genealogical approach as such, its aims may be useful to prehistory. The purpose of a genealogical investigation is not to ask for the "self-understanding" of the text (in our case, material), but instead to ask what the text (or material) itself does not "consciously" know is the subject. Genealogy is opposed to hermeneutics. The hermeneutic perspective attempts to understand the text in its own terms-to approach the text's own understanding of itself. The genealogical perspective poses wholly new and different questions. For a genealogical perspective, no questions are given or necessary or obligatory. The scholar should not work on the premises of the text but on her/his own premises. The scholar should herself formulate questions that cut across / oppose the text itself.

It may be suggested that such a perspective opens up much wider perspectives.


There are several systems linking "signs", "icons", "indexes" and "symbols". "Gestures" and "words" are frequently assumed to be symbolic. Regardless of the particular order of any given authority, there is usually a hierarchy in which a "symbol" has a greater "surplus of meaning" than a "sign".

How do these concepts come into being? At the most elementary level, one can check the OED.

In Greek the word meant only a "sign", "token", even a "ticket" or "license" according to the standard Oxford Dictionary of Greek (Liddell & Scott). The Oxford English Dictionary says however, that it is, i.a., "An object representing something sacred" or even "a formal authoritative statement of the religious belief of the Christian Church". A decisive change has taken place between the ancient and the modern definition, at least partially related to the appearance of the Christian Church and its specific use of symbolism.

This means that we have infused "meaning" into a term which we then oblige ourselves to define. More importantly, however, it also means that something which was eminently practical (a "ticket" or a "sign") has been assigned to a domain where it is assumed to have no "practical" value, at least by procedural definition if the "symbolic" is to be separated from the "practical".

The suggestion that among the criteria for a "symbol" must be the fact that it represents something related to belief is related to Christianity. In prehistoric archaeology things identified as representing "belief" are identified precisely because we cannot find any other purpose for them. However, the process of filling a symbol with meaning is precisely a historical process in the creation of meaning.


The difference between a Greek concept of "symbol" meaning a "sign" (cf. Liddell and Scott) and the definition given by the OED suggests a historical development of the term, from meaning something concrete and practical to meaning an untouchable thought. This development has certainly occurred in the last few thousand years. How much of the entire evolution of the human understanding of this type of symbolism has taken place since the beginning of the Bronze Age? And how much did the Neolithic bequeath? Was there any symbolism in the Palaeolithic?

Most agree-we do not-that sophisticated language can be traced back to the Upper Palaeolithic. This assumption effectively eliminates a gap in the development of expression during which people will not have had an "abstract" means of describing "tools" or "symbols", let alone distinguishing them. Before the development of any kind of language, conceptual thought will have been even more concrete and sexual since only compelling and memorable images could provide a means of categorizing the creations of the human hand.

It seems to us that in most definitions of symbols and in most treatments of symbols, be they archaeological or not, it is assumed that "symbolism" and its use is innate in mankind and that it has existed continuously the past 40,000 years. In other words, it is assumed as a matter of course that "symbols" existed and were used consciously in the Palaeolithic as well as in the Bronze Age, and in precisely the same manner. "Symbol", then, is assumed to be a constant, closely connected and almost identified with being human.

Contrary to this point of view, we want to suggest that "symbols" came into being only with the beginning of language, and that languages with abstract values may be quite recent. The result may have been a transformation of forms of expression which may have begun as little as 10.000 years ago. We therefore suggest that "symbols" existed only rudimentarily in the Palaeolithic (if at all) and that they can only be talked about as a partly "conscious" phenomenon from around the beginnings of the Neolithic.

The language of the material culture of the Neolithic is that of capacity and content: containers to be filled. Such materials lend themselves to metaphorical transformations. These metaphorical transformations employ and enhance language. This type of language-for expressing thought-is a common phenomenon today. But how old is it?

The capacity to express is visible in the material culture of the Neolithic and the languages of the Bronze Age. To what degree are we justified in making similar propositions about the Palaeolithic? What is the difference between "overwhelming potentiality" which cannot be expressed because there are no words for it, and a "surplus of meaning" which cannot be expressed because the number of associations is infinite, and thus inexpressible because of individual limitations and not because of the limits of language? Will this have no impact on the "meaning" of material objects?

The initial stage is the creation of a thing: a painting, an axe, a vessel, an ornament. A next stage is developing a "name" for that. It is only in the "final" stages that (a) objects become "otherworldly" symbols simply by virtue of names which were once descriptions, and (b) the process of creating objects which are consciously "symbolic", representing the real world in terms of the metaphor. In this fashion, a perfectly ordinary object, e.g., a jar, can be given a sexual significance through a name or suitable decoration. The reverse happens when, e,g, creating a statue of a woman and calling it "Aphrodite" so that the symbolic comes to represent "sexuality" and the world of being, but also much more.

The role of language in expressing the thought and allowing the transformation should not obscure the fundamental influence of the environment in determining the means of expression: the forms of reference remain the same. We can recognize a statue of Aphrodite as a symbol. We can also recognize a jar as a symbol. The issue is exactly at what point we can assert that the jar was a symbol. The concept of the woman as a "vessel" is reasonable. Does a jar decorated with a face and breasts "represent" a woman or does the character of the woman as a "vessel" fade into an inanimate jar?

Are metaphors symbols? Or what is the relationship between metaphor and symbol?

Gardiner (1963) stresses that language use is about discussing "things". Lakoff & Johnson suggest that human thought is to be understood through metaphor. We would prefer to argue, with Gardiner, that human language can only express relations of "things", and therefore that "metaphor" is the only possible means of expressing something which is not a thing. Applied to language, this may or may not be debateable, but applied to symbols there can be no doubt that these are "things". They are, however, "special things".

Many have followed Lakoff & Johnson, assuming that metaphors are a way of thinking, rather than arguing-as we do-that metaphors are a means of expressing a thought rather than a thought. And that this means of expressing thought is dependent upon language. The human mind processes images and uses these images to "describe". These "descriptions" are formed metaphorically in the absence of any other means to express the thought. The basis of the thought can frequently be an artefact, and the means of expression will be in terms of the environment. Language transforms this capacity, introducing an interface which can distort and organize impressions.

We find it to be crucial that in prehistoric archaeology we face the origins of symbolism. Archaeologists should therefore bear in mind that the symbols in the archaeological material do not necessarily look like symbols in modernity, and certainly not symbols as they are defined and discussed in textbooks. Can we assume that "symbols" (as well as religion) were "born" full-fledged as we know them today? Or rather should we-when looking at the archaeological record-look for incipient symbols, would-be-symbols, experiments of symbols, rudiments of symbols?

We suggest that, contrary to the means of defining and identifying symbols today (as signs referring to something else by way of association/ convention / as signs with a surplus of meaning), what we today see and interpret as prehistoric symbols, were in fact at that time not invested with any meaning in this sense. The history of the world is the "production of meaning" (Merleau-Ponty), and directly related to "history" in the sense of those divisions of human life which can be defined. As the origins of symbols and as experiments of symbols, these cannot have been more than mere signs with a potential, and that this potential was realized in the course of history, during which they were filled with "meaning".
Defining the "Criteria of Symbolicity" is thus actually another step in this same process by which incipient meaning is poured into ancient containers which were empty.


It is assumed that "symbols" have some "meaning" and "meaning" is usually associated with some "value" (even in the sense that mathematical expressions have a given "value", where the term "value" means simply a numerically expressed content). In the Graeco-Roman world it was considered incorrect to dedicate iron-rather than bronze-objects to the gods or the dead. In Hindu thought, a stone temple is tens of thousands of times more valuable than one of brick or wood. The intrinsic value of the object in terms of "price" was thus a component of the "value" of a "symbol": not the object itself, but the "price" transformed its "value".

In the Near East, there is a leap from the Neolithic to the Bronze Age. Up to the end of the earliest phase of the Neolithic, one finds objects of bone, decorations in colors, figurines in clay, etc. It is only during the Bronze Age that objects of intrinsic value-gold, lapis lazuli, etc.-began to dominate to "symbolic" world, whereas the intrinsically symbolic of the Neolithic was primarily symbolic through its character and not its material-clay, stone, etc. The offerings of earlier ages-boars' mandibles, dentalium necklaces-only became symbolic through context (burials). This again stresses that our approach to "symbolicity" must be grounded in a social and technological context.


We are looking for things which somehow became distinctly symbolic. We must also look for a reasonable time "when" symbols became "symbolic". We must also discuss our reasons for explaining "how" and "why" they became "symbolic"-even if we cannot answer, we must lay the thoughts on the table.
Our concept of symbol must be that we agree that a "symbol" must have a "surplus of meaning"-as opposed to an ordinary item. A symbol must be an expression of something (Saussure) and it must be recognizable (Peirce). It can achieve its symbolic state by any means. One debate in Semiotics opposes a Saussurian perspective and a Peircean one; the former arguing that the relation between a symbol and its meaning is not wholly arbitrary but contains some form of motivation (e.g. the goddess with a scarf over her eyes as symbol of justice) and the latter arguing that a "symbol" is characterized by the fact that it refers to its object only by convention and habit-in contrast to icons and indexes-(a classic example is the cross).

We cannot choose between Saussure or Peirce, because the distinction is illegitimate, the issue is not how the "surplus of meaning" came to be identified with an object, but that it did. Our problem is that it is almost impossible for us to identify any genre of symbol according to the meaning assigned to it before the appearance of language. Once we know that the goddess of justice is blindfolded because justice must be impartial, we can follow the system. Once we know that the son of god was allegedly executed by being nailed to a cross, we can follow the system. Otherwise a blindfolded woman with a sword and a balance might be a prostitute, and a cross might be a geometrical symbol generated by mathematics such as those of the Hindu mandala. Obviously the key point is the agreement, not the origin of the metaphor. In prehistory we cannot identify meaning.


Gardiner, A. H. 1963. The Theory of Speech and Language. Oxford: Oxford University Press.

Lakoff, G. 1990. Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago: University of Chicago Press.

Ricoeur, P. 1976. Interpretation Theory: Discourse and the Surplus of Meaning, Fort Worth.


Information: Paul Bouissac   
Design by: H. Harris
Copyright ©2003. All rights reserved