Legal document with seal impressions from Assur, Late Bronze Age. VAT09009. Credit: Staatliche Museen zu Berlin, Vorderasiatisches Museum / Sandra Steiß, CC BY-SA 4.0.
Front and back of a cuneiform tablet. On the back, there is also a rollout of a cylinder seal.

Controlled vocabularies

To talk about the same thing, we need to understand what the other person means. Even more so, if we aim to build a complex knowledge graph that integrates the full range of information accessible through seals and sealed artefacts. This includes the physical properties of the objects, their retrievable relations to space, time and people, but also the features that turn them into powerful agents in multimodal systems of communication–the pictorial and written signs they carry.

All terms used within the KIŠIB’s digital infrastructure are or will be defined in controlled vocabularies. These vocabularies provide the shared language through which objects, images, texts, and contexts are described, compared, and connected.

Wherever possible, KIŠIB’s vocabularies are mapped to established general or domain-specific authority files.  Specialised vocabularies developed by the project are published in SKOS format, making them available for reuse and cross-referencing by other initiatives. The vocabularies are continuously revised and expanded as the corpus grows to include new regions and periods, and they therefore remain a work in progress.

The following sections describe the main vocabulary areas used in the KIŠIB project, the kinds of information they capture, and the external reference systems to which they are linked.

Artefact properties and classification

Seals and sealed objects are physical artefacts with distinct material characteristics. KIŠIB records a range of properties for each object, including dimensions and weight, state of preservation, and details of material, colour, surface treatment, and texture. Where identifiable, manufacturing techniques are documented, as is evidence of repair, reworking, or later modification. Together, these properties allow researchers to trace patterns in processes of manufacture, adaptation, and usage, and thus the ‘bibliographies’ of objects across time and space.

Ancient West Asia produced a wide variety of seal types and sealed artefact forms. KIŠIB classifies these objects along three main dimensions: the type of seal (e.g., cylinder seal, stamp seal, or a figural stamp seal like a scarab), the type of sealed artefact that bears an impression (e.g., clay tablet, clay envelope, hollow clay ball), and the type of object to which a sealing was originally attached (e.g., a basket, a vessel, or a door bolt). These distinctions are important because the form and function of sealed objects are closely tied to specific administrative, legal, and economic practices. A clay lump sealing a storeroom door served a different purpose than a seal impression on a legal contract, and the classification of these forms helps to reconstruct the social contexts in which seals were used.

Pictorial entities

One of KIŠIB’s core objectives is the systematic annotation of figures and details depicted on seals. This is achieved by connecting selected portions of a seal’s image—marked by bounding boxes—with terms defined in a controlled vocabulary. The vocabulary is structured hierarchically, moving from general descriptions of a figurative composition as a whole (thematic group) to the identification and detailed characterisation of individual pictorial elements (picture element class). The aim is a transparent and consistent labelling process that avoids synonyms and terminological inconsistencies for the same figure or feature. Since the corpus is continuously expanding to cover new regions and periods, the vocabulary will grow accordingly and remain a work in progress. Rather than imposing a single standardised terminology, KIŠIB aims to document and connect the diverse naming conventions that have developed across more than a century of research in different languages and disciplinary traditions. The goal is not to flatten this diversity but to make it transparent and navigable—preserving the intellectual history of seal studies while enabling structured comparison across the corpus. Wherever possible, the KIŠIB picture vocabulary is mapped to existing authority files for art and iconography, especially the Getty Art & Architecture Thesaurus (AAT) and Iconclass.

Picture elements are characterised by their pictorial rank—distinguishing main figures from secondary and interspersed elements—and by their identification (entityclass), that is, what or who is depicted. For anthropomorphic and zoomorphic figures, the vocabulary records sex/gender if marked through distinctive visual attributes, body and face orientation, posture, and gesture. Personal attributes, such as objects held, carried or used by a figure, are documented along with their position relative to the body. Finally, a detailed set of descriptive categories covers elements of appearance: beard and hairstyle, headdress, clothing and textiles, jewellery, and footwear. Where written elements appear within the image field, the graphic inscription layout is also recorded.

Artistic style

Stylistic classification is one of the most contested and complex aspects of art-historical research, and seal studies make no exception. Different scholars have approached the question of style from various angles, often leading to overlapping or conflicting categories. Style designations naturally intersect with spatial and temporal classifications, yet the boundaries and connections between them remain debated. A seal described as belonging to a particular “style” may carry implicit assumptions about where and when it was produced, by whom, with what message and for what purpose—assumptions that are not always made explicit.

KIŠIB addresses this challenge by treating stylistic classifications not as fixed categories but as research-historical constructs. The definition and attribution of style are never neutral but always depend on scholarly decision-making, research agendas, and, often, implicit assumptions about social practice and material culture in antiquity. KIŠIB therefore consistently links each style designation to the individual scholar and publication from which it derives, rather than hardening these attributions into authoritative categories. This approach fosters multiperspectivity and ensures that users can trace, compare, and critically evaluate competing classifications in the context of their own research.

The vocabulary for artistic style will be developed incrementally as the corpus grows and as stylistic patterns become amenable to systematic analysis. Rather than aiming for an authoritative taxonomy, KIŠIB treats artistic style as an entangled history of research—a domain of collaborative, long-term refinement in which competing perspectives are documented and connected, rather than resolved.

Spatial entities

Seals and sealed artefacts are linked to places in multiple ways. They may have been archaeologically excavated at a known site (place of discovery), acquired from a dealer or collection with a stated (but not always reliable) provenance (place of acquisition), attributed stylistically to a certain region (place of production) or identified through textual evidence as used by a person or during an event, for whom or which we posses geographic information (person: place of residence; event: place of usage). Today, they are kept in museums and collections (current location). KIŠIB’s spatial vocabulary must therefore accommodate both geographically fixed locations and historically attested place names, and it must be able to express from where information about spatial attribution derives and that varying degrees of certainty for that attribution exist.

At the most concrete level, the vocabulary records archaeological sites and modern places, each defined by coordinates. Sites and places are assigned to regions and subregions, following a regional framework based on the ARCANE initiative, with supplements adapted to the specific needs of KIŠIB. Wherever possible, spatial entities are mapped to established authority files and gazetteers, including GeoNames, Pleiades, the Getty Thesaurus of Geographic Names (TGN), iDAI.gazetteer, the GND (Gemeinsame Normdatei), the CDLI Region Index, and the CDLI Proveniences Index.

Temporal entities

The chronological classification of seals and sealings presents challenges similar to those encountered with spatial attribution. Dates are rarely absolute; in most cases, the temporal assignment of an artefact rests on a combination of archaeological stratigraphy, stylistic attribution, typological comparisons, or associated textual evidence. KIŠIB records temporal information on various points in the life of the object (times of production, usage, modification, discovery, acquisition) and at multiple levels of granularity, from broad period designations (e.g., Early Dynastic, Old Babylonian, Late Bronze Age, Achaemenid period) to more specific sub-periods (Iron Age IIIa) and, where available, absolute dates or timespans. The vocabulary for broader temporal categories aligns with established periodisation schemes in ancient West Asian studies, while remaining flexible enough to accommodate regional variation and ongoing revisions to the chronological framework. Temporal entities are mapped to PeriodO, iDAI.ChronOntology, the CDLI Periods Index, and the GND. As with spatial entities, KIŠIB employs a differentiated system to document the basis and degree of certainty behind any chronological classification.

Personal entities

Seals are deeply personal objects. They were carved for individuals, inscribed with names and titles, and used to represent their owners in legal and administrative transactions. The study of seals, therefore, requires a robust system of identification for the persons associated with them—both those who lived in the ancient world and those who have studied, excavated, collected, or published them in modern times.

KIŠIB distinguishes between several categories of personal entities. Modern persons include authors of publications, excavators, and project contributors. Ancient human persons encompass seal owners, officials, and rulers. For each individual, the project differentiates between personal names and historically identified persons, since the same name may refer to different individuals across periods and regions. A further category comprises divine and supernatural persons, including gods, goddesses, and hybrid beings, who are frequently depicted on seals and named in seal inscriptions. Their identification is complicated by different writing conventions, regional variations in divine names, and phenomena such as syncretism, where distinct deities merge or are equated over time. KIŠIB records these complexities rather than resolving them into a single canonical form. The project’s person lists are built on datasets compiled from the corpus itself and cross-referenced with established prosopographic resources, including data from the Open Richly Annotated Cuneiform Corpus (Oracc). Personal entities are currently mapped to the GND and Wikidata, with further mappings to other resources planned as the corpus expands.

Social entities

Beyond the individuals who owned or used seals, the corpus also captures information about the social roles, institutions, and organisational structures in which sealing practices were embedded. Social entities in KIŠIB encompass the titles and professions attested in seal inscriptions and associated texts, the social roles that individuals occupied in relation to specific administrative or ceremonial practices, and the institutions—palaces, temples, households—within which seals were produced, stored, and used. This vocabulary is closely related to both the personal and temporal entities described above, since titles and institutional affiliations changed over time and varied across regions. By treating social entities as a distinct category, KIŠIB enables queries that cut across individual biographies—for example, tracing how particular administrative roles were associated with specific seal types or iconographic programmes, and with officials employed in the same institution.