A TRIPLE SEMIOSIS FOR THE REPRESENTATION OF STONE TOOLS IN ARCHAEOLOGY
(Paper presented for the Fifth Congress of the International Association for Semiotic Studies, Berkeley, June 12-18, 1994)
Giovanna M. Winchkler
Abstract: This paper aims at showing the possibility of the use of Peirce's model, as a general orientation that would allow the enrichment of the different levels of this language by finding in it the three kinds of operations (comparison/similarity, operative inference, and valuation), that the archaeologist carries out when using the language and stating the assumptions that make it effective.
The archaeologist's mind implements different kinds of operations when he/she uses the texts which are at his/her disposal to know how to observe. Taken as real assumptions of observation, the three kinds of relations with which Peirce organizes his system of signs (Peirce, 1965; Magariņos de Morentin, 1983), help to describe such operations with sufficient rigor.
Archaeological discourses are discourses about other things (that which is spoken about), as any discourse and, thus, it can be organized by means of Peirce's three main types of signs, a kind of syntax considering a particular association of such signs.
On the other hand, when studying some discourse (whatever its quality may be) about another thing, two levels of signs are being dealt with: the sign which makes reference to another thing and the sign which is that thing another sign refers to. In this sense, the signs which refer to another thing (substituting semiosis, Magariņos de Morentin, 1990), can be studied as being icons, indexes and symbols, and likewise, those which are the thing the other refers to (substituted semiosis, id.), can also be studied as being either icon, index or symbol. Thus, the quality of sign of the representamen and the quality of sign of the ground can also be taken into account.
Peirce's cognitive operations in archaeology
The cognitive operations the archaeologist performs when going to analize the
stone tools, can be classified, following Peirce's scheme, as follows:
1) organizing a mental image that would allow to look for its similar one in the stone tool;
2) organizing the representation of an activity that allows to infer or to see the results of its presence in the stone tool;
3) representing a system of relations that would allow the identification of such relations in the stone tool by measuring or designating it (that is, including what can be seen in a paradigm of related terms).
The operations that observation implies could be then:
1) the similarity of possible forms;
2) the behaviours inferred by the concrete configuration of existing elements;
3) the intellectual behaviours that look for correspondences among abstract concepts.
The archaeologist can organize a cognitive proposal:
1) from a graphic representation, as those drawings through which the contours of certain elements are outlined. Then she/he can find out if the shapes observed in the stone tool are similar to the drawing, for example, of a triangle. But this can also be carried out from the observation of a gesture, such as by outlining a triangle in the air or by reading the written expression 'triangular';
2) from the representation of an activity that is supposed to affect the stone tool leaving marks on it. For example, a session of experimental flaking configures the action where the events are contiguous and of effective realization, but they are representing other actions and events not present;
3) from relations which have no presence in the world of concrete stone tools nor the density of an imagined shape, but that perform certain abstract operations that assign values, names, labels, places in a system, like, for example, measurements, the setting of relations between axes (generally, more abstract than real), the classification of shapes of not perceptually evident sections, the ranging in conventional hierarchies, etc.
The operations that the observation of the stone tools implies could be, then, of the three types
proposed by Peirce:
1) similarity: icons;
2) association by contiguity: indexes;
3) intellectual operations by convention: symbols.
Semiosis in archaeology
As regards the material part of the sign, it can be differenciated:
1) the substituting semiosis or representamen, as the sign which represents another sign, that is to say, that part of semiosis constituted by the reproduction of objects into drawings that represent the shapes of the substituted sign or ground, through the outlining of contours and spaces. Consequently, it is iconic;
2) a part constituted by operative chains and sequences of use (that includes technical or functional actions or gestures and objects that complete them) or of deterioration contextually linked to shapes produced by these activities. Consequently, it is indexical;
3) a language or ensemble of conventional signs constituted by the texts, therefore, symbolic. It is written text, natural language while it is not formalized. But it is cientifically used, since its terms are specific of archaeology in as much as they are redefined in archaeological texts (Winchkler, 1993). Because of this cientific usage, it can be considered as the specific language of the correspondent domain, which in this case is lithic analysis.
Bounding the sign
In those material presences, where is the sign? Each single word, out of context, is it a sign? Is an expression a sign? Or is a complete phrase or definition a sign?, etc. (Peirce (2.230) says that a sign can have more than one object and he gives as an example a brief phrase; some other times he gives as example of a sign some complex phrase and, sometimes, individual terms).
As regards to the problem of drawings, archaeologists do not generally present them in their writings as elements independent of the written part but as complementary ones, so drawings are not a parallel language. And in fact, drawings almost always add something to their correspondent text, but with a varied relation to it. However, in some cases they are, in fact, parallel to the written text, for instance, when the written part consists of a mere enumeration of words referring to geometrical shapes. The list of drawings representing those shapes, under the same title of the written classification, could be said to have the same ground (it refers to the same aspects of the same things) though the material support or mode of presentation varies (word/drawing). Neither the word clarifies the meaning of each drawing nor, conversely, do the drawings exhaust the meaning of the words. They are not the same sign, though they may refer to the same referent.
A triangle drawn at the side of a text which enumerates the shapes of triangular projectile points, is not the same kind of sign as a small triangle joined by a dotted line in one of its edges to the lower edge of the drawing of a contour. This small triangle, which may be not referred to in the text, shows, by convention, the cross section of the object depicted, for instance, a flake.
Now, if the text makes specific reference to the triangular shape of the transversal cross section of the flake, the same previous figure has another sense: it only shows, as an added information, the location in the flake, of the element the text describes. Also, if we take this last drawings out of the context of the written text, there appears a sign that is different from the two which complement each other and to the text, since this new one shows a totality (the flake and its transversal cross-section), instead of laying emphasis in the shape of a certain cross-section or its relative location in the flake. This means that restricting a sign may also require certain criteria.
I consider that the study of the sign in archaeology can be made through different approaches, depending on the position one wishes to adopt to study the material shapes the sign takes (representamen): drawings (the iconic), gestures (expressed live, photographed or filmed, that constitute different indexical signs), texts (the symbolic).
The language of drawings or iconic semiosis
Let's go back to drawings: those which most frequently appear as complementing terminologies1 are, sometimes, like a zoom that illuminates the aspect that is being described, like an extrapolation of some part of the stone artefact. For example: a sharp-pointed part of a tool can be morphologically differentiated according to its relation to the rest of the tool is of 'separation' or of 'connection'; it also can be differentiated by juxtaposed edges or by alternating edges in a situational relation as regards to each face of the artefact. In these cases, the drawing tries to magnify details that, in a global image of the object, could make more difficult the operations that help to recognize the aspect of the tool that the text wants to recover.
The text (let's consider the symbolic quality of any language as accepted) says something of some other thing. Drawings (leaving aside their quality, not always necessarily iconic) are also substituting something. But, which is the semiotic object created by each one of the two languages in this particular case? And, since both kinds of language are often used in mutual relation in terminologies, specially in taxonomic ones, how do descriptive phrases and drawings relate to each other as being signs and cognitive operations?
The text, in the former case of the example, refers to a relation (connection/separation of body vs. rest of the piece) and, in the latter, to another kind of relation (yuxtaposition and alternation of edges on the piece's faces).
But do the drawings express connection/separation, yuxtaposition/alternation? Not whithout the text. In isolation, any archaeologist can interpret them as representations of parts of pieces, but the specific relations expressed by the text cannot be recovered only by means of the drawings. No semiosis (or kind of sign) is sufficient for explaining itself; its interpretation needs the information resulting from one of the three possible kinds of semiosis. Drawings complement the text: they do not refer to the object as aspect of the world, but to the object as aspect refered to by the written text in a certain way. The text recovers its capacity to signify thanks to them: the reference of the text is, in each case, the way two elements in the piece mutually relate, but the drawing's reference is, in each case, the iconic shape, that observation is to seek, not as it is shown by the bareness of the drawing, but adding what the text says. The semiotic object is constituted by two kinds of languages (writing and drawing) integrated in only one semiotic syntax. As regards to the substituted semiosis: in the written text, what is described is the relation connection/separation, yuxtaposition/alternation, that makes it possible to differentiate both sharp-pointed forms. What is represented in drawings is interpretable as part of the piece. It is not a relation what is substituted in them but a sharp-pointed portion of a piece. (This is not the place to answer why I can say that the drawing is archaeologically 'interpretable' as part of a piece, as a sharp-pointed portion. The mentioned drawing consists, for 'natural' perception (vs. the archaeological one, since there is a scientific use of language that constructs objects which are different from the ones natural language allows to construct) of two curved lines, with different curvature, that intersect in acute angle, with a series of insertions drawn along the lines. So, it would be possible to talk of a symbolic level that preexists the act of interpretation through which we, archaeologists, understand each other iconically).
Among the drawings of frequent occurrence in archaeological terminologies, there are those that constitute very schematic representations of geometric shapes, as, for example, section's shapes. The geometric names of those shapes are detailed through the yuxtaposition of a phrase and a drawing; moreover, sometimes there are composed by a word and a drawing. It is not necessary to be an archaeologist to find a meaning to parallellogrammic, trapezoid, triangular, or the like. The constructed object is the same in both languages and together, they don't change such interpretation. What changes is the quality of the sign: it would be a symbolic sign in the case of the text and a symbolic icon in the case of the drawing. In the former, the archaeologist conceptualizes a shape he knows by convention, in the latter, he checks that the shape he observes in the object corresponds to the drawing. It could also be studied: a) if a series of stereotyped geometrical shapes, drawn in orderly arranged sequence, isn't as conventional as the language in which such shapes are enumerated; b) the additional assumptions needed for the correspondence of the shape observed in the tridimensional object and the drawn shape (Marr, 1982).
Other drawings of these terminologies represent schematizations, but not of geometric shapes. For example, in order to indicate the inclination angle of the butt in the ventral face, the outline of the piece must be drawn in such a perspective (of the blank and of the flake) that it shows this angular difference. Each drawing has sense because the others are there; in itself, in as much it goes together with a text that is about measuring angles and expresses possible values, it would be as the previously mentioned case, of the stated and drawn geometric shapes. The series of drawings visualizes the butt's change or variation as a consequence of that angle. In the case of the percussion bulb, the role of change indicator that the drawing assumes is stressed because (with the exception of a few cases) there is no description of what is understood as 'soft' and 'marked', which are two of the categories of bulbs. Other categories of the same classification of bulbs, as 'non-perceptible' and 'non-determinable' do not obviously correlate with drawings (in the second of these, according to the texts, the observation cannot be made due to fracture, disintegration, etc., it seems not to be representable. But then, what do they correlate with?).
In certain cases there are series, represented as if they were composed by photograms, refering to the sequence of production of a tool: the series would be an index (indexical icon, because there are drawings; if they were photograms or in video, it would be index, according to Peirce, but as regards to its graphic quality, it would share the icon's nature. In addition to being icons for their shape or representamen, drawings also reproduce other shape).
This enumeration doesn't exhaust the ways drawings and texts appear in the terminology. The outlines of the pieces are drawn on a grid with squares differenciated by sectors. Being twice drawn, the image complements itself with a text which simply expresses the names of the faces, number of sectors, etc. The shape of the piece itself does not matter here; what matters is the relation between the piece and the other element, the grid squares. It is a symbolic icon (since it's a drawing), it shows a convention. The same happens when two intersecting lines are simply represented by means of a semi-circle showing, for instance, the angle of a barb or an edge or some other element. This sign would also be a symbolic icon. But, on the one hand, it is marking a convention not pertaining to archaeology, while the one of the grid squares represent the specific way of placing the pieces in archaeology, and on the other hand, it relates two icons to construct that symbol (the one of the grid squares, whose referent is a possible way of distributing and organizing the space (the one of the grid squares as ground) and the piece, whose referent is the concrete piece placed on the drawn grid squares).
The language of gestures or indexical semiosis
When a person imitates the gestures of another (as for example in the workshops where flaking is learned) his gestures would be iconic gestures. But, when replicating the operating chains, by means of which, we assume that the craftsman of the past obtained the tool we are reproducing, our gestures would be indexical indexes because, given that assumption, both series of gestures are, in some way or other, in the same context, being the gestures of the past necessary for replication to exist. On the contrary, for example, to express the angle of a butt, it is habitual for the archaeologist to make a gesture downwards, towards the tool he his holding, tracing a line in the air. It can be understood immediately as indicative of certain angle of percussion. Such gesture could be a symbol of the percussion angle of the butt, that is, a symbolic index.
The written language or symbolic semiosis
When the meaning of what is said in the text depends on a comparison operation between two semiosis (being usually words or images), we are dealing with an iconic symbol (for example, that shapes of which we have an image in the natural language: 'on top', 'cornice-like', 'egg-shaped', etc., they would be iconic symbols); when the meaning of what is said in the text requires the contextualization of the expression by means of gestures or actions or with some other words, we are dealing with an indexical symbol. Such as, for example, 'burin blow fracture', 'impact point', expressions that denominate entities and transport the idea of the action or actions carried out to obtain them. These signs require the realization of certain operative inferences (different from the inferences required by the interpretation of other signs and related to the operation that Peirce attributes to the index: the association by contiguity). Many names we use in the stone analysis are of that sort, that is, indexical symbols, since they indicate processes and actions whose traces are observable through their different states, which, in general, are icons. When the meaning of what has been said appears as a convention, it is a symbol.
Diagram of semiotic possibilities
Considering drawings as icons (as being signs that reproduce the shape of things), we have these different kinds of signs:
-Cualisigns (as ellaborations of qualities or possible forms of existence): textures or contrasts, linearity or photographic dashed surfaces, with which the objects of the world are being represented.
-Icons as such (as they represent concrete shapes of existence): the drawings that represent existing characteristics, such as: cornice shaped, peak shaped, amygdaloid (that is, amygdala shaped), etc.
-Rhemas (that contain the value of shape): the drawing of a small flake ('microlasquita') next to the negative mark of flaking of the corresponding flake. It would be only the value of the shape, if we wanted to show the negative/positive complementarity, but not if we wanted to recover the employed technique. In this case, we use the drawing to show a possible operative inference. It is rhema only while it is constituted by image systems or shape sequences representing the possible variations of graphic representation in archaeology.
Considering gestures or actions live, photographed or filmed, as indexes (in as much as they are signs that substitute existence in its proper act of existing), we have the following signs:
-Sinsigns (as they are representation of events, objects and behaviours captured in the phenomenic specific characteristics of their existence): the individual gestures which constitute a complex knapping behaviour (putting the anvil on the soil in a given position, sustaining a blank in one hand, sustaining the hammer in the other hand, etc.).
-Indexes as such (as they are existing actions which are signs since they represent other existing characteristics, of which they are a part that cannot be separated from): the integrated sequences of gestures in the context of a workshop where knappers experiment flaking. The operating chains that the knapper's actions perform represent the actions of the craftsman of the past, whose effect on the stone knappers try to replicate.
-Dicisigns (as they constitutes the context of indexes): the knapper's action and its imprints in the tool, that are integrated in the same but more ample sign. This inference, that I am calling operative to differentiate it from the one that relates every sign with its object, links -in a broad sense- gestures and traces in archaeology.
Considering written texts as symbols (in as much as they constitute an aspect external to the object but inherent in it, due to the significance that a culture -and, in this case, a scientific community- attributes by convention to it), we have the following signs:
-Legisigns (as they do not have value for themselves but for being normative features destinated to produce symbols): in the context of a terminology, for example, the syntactic grammatical characteristics conferred to them by their quality of being names, adjectives, verbs, prepositions, etc.
-Symbols as such (as a language carrying specific contents): the specific language of archaeology, that accounts for the stone objects -which is the one I am studying in this paper.
-Arguments (as they are representations of the values that obtain in the scientific community at a given moment and place): the formal or systematic aspects of the language with which we study the stone tools. For example, their degree of formalization or the gradients of freedom or vicinity with the natural language postulated for it, from the more ample scope of archaeology. In fact, it is constituted by the set of cognitive operations, explicit definitions, terminological options, and rules for interpreting, that constitute the specific difference between the theoretical trends relative to the usage of archaeological language.
Each sign establishes, due to its capacity of representation, a cognitive operation that, within semiotics, has certain features. More than applying Peirce's categorical model, I consider fruitful to use Peirce's signs, as a description of the cognitive operation that such signs are performing rather than as the category places of a taxonomy. It means exploring the usefulness of the concrete bases which Peirce offers for the enhancement of the objectives of this research about the visual, behavioural, or verbal representations of the lithic tools in archaeology.
1 I consider as terminology (Hartmann and James, 1998) those texts which make use of the terms of lithic analysis for defining them, or for describing artefacts or items of classifications, etc.
References to this paper
References to the archaeological corpus of the broader research