A Grammar of the Archaeological Record (Version 2, Beta release)

Epistemics. Acquiring knowledge

Structuring: wholes

Giorgio Buccellati – October 2025

From fragments to wholes

The disentangling of the matrix, whch we have seen to be the hallmark of the stratigraphic process, extracts elements which are fragments of larger wholes, wholes that are not immediately apparent in the stratigraphic record. What is required next is a structuring process that identifies these wholes.

We deal here with a semiotic dimension, one that aims to replicate the semiotics of the ancients on the basis of objective patterned correlations of formally defined (hence, grammatical) traits. The ancients dealt with wholes, not with fragments – the grammar guides us in identifying these meaningful wholes, by allowing us to establish patterned correlations among formal traits.

Types of “wholeness”

There are different types of “wholes,” and the images below provide two examples.

An aggregate: burial A8a8
An assemblage: 23 selected conical cups

The tomb on the left is a whole as it is found in the ground: a coherent structure, partly damaged (the vertical bricks that formed the top have fallen in), but with its original content still in place: the brick structure, a skeleton, individual offerings, accumulations that derive from the collapse of the roof.

The image on the right represent a whole as restructured by us, a collection of similarly shaped items, which we call conical cups. These particular twenty-three cups were never seen together in antiquity: they have been put together by us from a variety of different places and strata, but we can be confident that they would be recognized as a coherent “collection” by an ancient in the same way we do – because of the objective nature of the formal traits they share.

Aggregates and assemblages

The main difference between tomb and cups rests on how each relates to place, and in this regard we distinguish between the two by defining one as an aggregate and the other as an assemblage (see already in the section on constituents).

Aggregate. – The tomb is a stationaty element, meaning that its identity is tied to the place where it is found. Qua aggregate, the tomb is present to us in the same way it was to the ancients: it comes to us already “structured,” since we recognize the aggregate semiotically as a tomb. It is an original whole.
Assemblage. – The conical cups are movable items, meaning that their identity is not tied to place. While this particualr collection never existed in antiquity, it is plausible to assume that it would be recognized as a whole by an ancient the same way we do. It is a whole which we have structured, and which can be said to have a legitimate semiotic valence. In other words, an assemblage is a derived whole.

Nesting

There is one additional level of complexity, because these wholes may be nested in progressive ranges of inclusiveness.

In the case of the tomb shown here as as example, it may be included in two parallel sets of elements.

On the one hand, it is to be seen within the wider context of the area where is was found – which turns out to be a sort of city of the dead, with a number of other burials in the midst of a living quarter of the city.
On the other, it should be considered as part of an assemblage of similar burial structures, from Urkesh and elsewhere, on the basis of a wider typology that relates to the shape of the tomb, the nature of the offerings, etc.

The conical cups assemblage, in turn, must be seen as part of a wider typological assemblage, in two regards.

Conical cups may be subdividede into finer sub-categories, depending on details of the shape.
They must in turn be seen within the full inventory of ceramic shapes from Urkesh, which at this point amounts to 936 types.

a whole which we may need to “disaggregate” in the course of the excavation, by dismantling the bricks and removing the contents. In other words, an aggregate is

palace

There are different levels of disaggregation. Both an object and an installation may be found whole or shattered and incomplete. And each may in turn be correlated to larger groupings of elements that match the same formal and functional traits.

The tomb then emerges as an aggregate, i. e., as a functional whole. But it needs in turn to be correlated to other elements

We may say that in this stage we aggregate the disaggregation. Through the definition of formal traits, we can recognize assemblages of structures and items that share the same morphology.

What results is a structured universe, which can, to some extent, claim to reflect the semiotic universe of the ancients. Given the patterned regularity of certain “classes” of elements, we may reasonably infer that they were perceived as such, i. e., as classes, by the ancients in a way that parallels our own perception of the same classes – as in the following example.

This patterned regularity emerges from extablishing correlations among formal traits of the data found at the site in the first place, and then correlations with different types of data and with with other sites.

The grammar

The grammar gives us the tools for translating this complex physical record into a "known" referential record.
Below we will briefly describe the nature of the grammatical approach.

PROCESS CONTEXT METHOD
2.
Acquiring
knowledge Structuring wholes Typology
Integrative

Process

The structuring process consists in identifying structural wholes

is based on

the identification of formal traits that are shared by a group of elements, which results in the definition of given assemblages, and
the correlation of these assemblages with data .

The validity of the process depends on the quality of the formal system of definitions, and on the accuracy with which the system is applied.

Context

What results is a structural whole, an assemblage, which may be further defined as

Aggregates

fixed/stationary structures,

Assemblages

either a “type” when viewed in terms of the formal characteristics that are shared by the individual elements,
or a “collection” when viewed in terms of the actual total of the elements.

Both stationary features (such as walls) and movable items (objects and samples) are seen in their identity as individual elements, but they are correlated to other elements of the same kind: they constitute an assemblage, the validity of which depends on the size of the inventory.

In principle, the validity of the context is independent from the total number of elements that can be identified, but in effect it increases in the measure in which the total number of elements increases. A type that includes only one exemplar is

Method

The essence of the structuring method lies in the clustering of the data according to intra-referential formal traits and extra-referential traits

Typology

The individual elements within an assemblage are seen to share formal traits, so that they form classes of elements. The notion of structuring refers to these classes, which are structural entities independent of their stratigraphic location, but grounded in the objective attributes that make it proper epistemic reality.

Typology – Once excavated, data can be assembled into meaningful wholes according to two distinct criteria. – Morphology ?

The first criterion looks at data depending on their intrinsic qualities: we construct typologies on the basis of inner-referential attributes, i. e., attributes that refer exclusively to the data as such, e. g., shape or material for ceramics, iconography for glyptics, paleography or linguistic analysis for texts.

Integrative

Integrative – The second criterion looks at the data with a view to integrate them into a broader picture, in terms of a variety of extra-referental attributes: these include comparison with data from other excavated sites; analysis of materials with techniques such as Carbon 14 analysis; confrontation with the broader historical framework as defined by textual data. – Syntax ?

nesting

The chart below gives the statistics of a larger selection from eight excavation units, giving the total of vessels and sherds and the totals and percent of conical cups.

unit	total	c.cups	%
A12	47,815	790	1.65
A15	62,117	552	0.88
A16	59,818	350	0.58
J1	44,732	262	0.58
J2	33,582	246	0.73
J3	13,301	13	0.10
J5	18,309	85	0.46
J6	17,182	80	0.46
total	296,856	2,370	0.80

Semiotics

see referentiality

A12: cc second highest [percentage after jars – https://urkesh.org/MZ/A/A12/D/X/XF/FAM-CA.htm