All museums and repositories use a numbering system to assign unique identifiers to specimens and groups of specimens, but with diverse and sometimes bewildering systems of notation. Accession numbers and catalog numbers are common and used to track different types of information. Accession numbers usually, but not always, refer to groups of objects. An accession is the transfer of ownership from an outside party to the museum/repository. Accession numbers are assigned to record the details of transactions by which objects are added to a collection. Catalog numbers often, but not always, refer to individual specimens. Catalog numbers are assigned to specimens because they are used to incorporate object metadata and record actions taken by the collection manager specific to those specimens.
Readers unfamiliar with museum curation should note that these terms "accession numbers" and "catalog numbers" have generally agreed upon meanings readily available in textbooks and manuals. In practice, however, the terms are sometimes used incorrectly, inconsistently, or interchangeably. This is especially likely when curation is being undertaken by personnel (e.g. archaeologists) whom are not specifically trained in museum registration or when established local conventions (i.e. "house rules") have dictated an atypical usage.
As with other types of museum collections, an accession number is often the first number attached to an incoming specimen or collection. Accession numbers are used in museums to denote a single transaction between a source (e.g. a donor) and the institution. An accession number often refers to a "lot" of objects, but it is also possible for individual specimens to be assigned an accession number if there is only one object in the transaction.
There are numerous systems in part throughout the museum community. For the purposes of this white paper, we focus on the two most common systems for accession numbers:
It is not uncommon to see additional numbers or letters used as prefixes and suffixes to these numbers and every institution will be different. Even a museum that follows strict conventions is likely to have many deviations that resulted from accessioning or cataloging activities undertaken in the institution's past. Most museums have multiple systems in active use, because they opt to continue the use of old existing systems rather than attempt to renumber or change an established system.
Most large museums use a double or triple-number system. For example, a natural history museum receives a single donation that includes two projectile points, a fossil, and a set of preserved leaves. In this case, this lot of objects would receive an accession number (in double-number format) which will be permanently shared by all of the objects. The accession number ("2016.10" denotes the tenth accession of the year 2016) exists to maintain the link between the objects and the transaction. The accession number will be accompanied by data about the donor, his contact information, the date of the donation, and other relevant information about the transaction itself. After an accession number is assigned, each object is then prepared to be sent to a different collection: projectile points to archaeology; the fossil to paleontology; and preserved leaves to the herbarium.
At this stage, each object can now be assigned a catalog number to distinguish it as an individual specimen which has its own metadata. In this case, the museum uses a two-number system to denote accessions and a three-number system to denote catalog numbers. The projectile points are assigned "2016.10.1" and "2016.10.2" and each has an independent catalog record that lists its provenience, point type, measurements, etc. The fossil is assigned the catalog number "2016.10.3" and its record includes different information such as species, geological age, etc. The leaves in the herbarium are assigned collectively to be "2016.10.4" because they are fifty identical leaves from the same tree and the botanist has no reason to treat them as individual objects. Should the botanist later change her mind, the catalog number could be continued (e.g. "2016.10.5"; "2016.10.6"; etc.); further divided (e.g. "2016.10.4.1"; "2016.10.4.2"; etc.); or suffixed (e.g. "2016.10.4a"; "2016.10.4b"; etc.) following institutional convention.
Two and three-number systems are popular in museum curation and are preferred by most large institutions. One reason is that a sequential single-number system is poorly suited for an institution that processes multiple new accessions at the same time. A sequential single-number system would require that each accession be processed one at a time in sequence before the next could begin. As a result, single-number systems are more likely to be used in small collections where there is little growth and few accessions per year. For small organizations, accession and catalog numbers may be one in the same and known by either name.
Depending on the institution, the systems of accession and cataloging notation may be entirely independent of each other or there may be an obvious numerical link between them, as in the examples given above. Many variations and other systems are also in use, but most museums and repositories tend to at least process new accessions by these notation standards.
Archaeologists routinely use multiple complex numbering systems for a wide range of purposes and there is virtually no standardization within the discipline. In particular, they frequently use arbitrary systems to number or otherwise denote provenience in a wide variety of systems. There are perhaps only two common numbering conventions that archaeologists routinely follow: site numbers and field specimen numbers.
Site numbers are a unique combination of letters and numbers permanently assigned to an archaeological site by a State Historic Preservation Office or equivalent. The system used is known as the "Smithsonian trinomial" or "Smithsonian tripartite" and has been used continuously since the 1930s. With some minor variations, it is used in all 50 states. The first two numbers are a code to indicate the state, the second sequence indicates the county or county-equivalent, and the final number is a running sequence which denotes the order in which a site was listed. The Lichliter site is "33 My 23" ("33" = Ohio; "My" = Montgomery County; "23" = Lichliter was the 23rd site listed in Montgomery County, Ohio). Site numbers are used heavily and usually appear on every bag, form, and physical source of data from a site.
The only other widespread convention is the use of a "field specimen" or "FS" number. An FS number is usually the unique identifier for an individual bag as objects are collected in the field and placed in the bag. FS numbers can take any format - their presence on an artifact field bag is the only reliable constant. When a sample is taken from a particular provenience, all artifacts and ecofacts are usually placed in the same bag. The FS number is the link between the bag and the provenience, which is itself a link to the context and associations of the objects. FS numbers always accompany artifacts and specimens throughout initial processing stages, but may or may not continue to be used in post-processing analysis or storage.
After a field bag is completed, the bag will then be sent to a laboratory where the contents will often be cleaned (e.g. washed or dry-brushed) since they are likely to be dirty from field recovery. Many other possible steps subsequently happen at this stage. For example, some objects may be re-examined after washing, determined insignificant (e.g. random natural stones picked up by accident), and discarded in the lab. After washing, other objects may be recognized as more significant than originally thought (e.g. a rim sherd that was not recognized in the field) and separated out as special finds. Some institutions will physically label the artifacts, while others will only label the bags or containers. Labeling archaeological specimens is notoriously problematic since artifacts and ecofacts can be very small.
Once a field bag has entered a lab for processing, there is usually one significant constant across all institutions. The materials are sorted into their different material types (e.g. lithics, ceramics), rehoused, and permanently separated. Material class is the primary criteria by which archaeological materials are divided because each material class will be examined by a different specialist. Special finds (e.g. tools, ornaments) are also likely to be isolated from the rest of their material class, if they were not already isolated in the field.
There is great inconsistency between institutions at this stage in regards to the FS number and the disposition of the now-sorted material types. The original bag is presumably empty and ready to be discarded, because each material type has been placed into its own independent bag. Any special finds may also be isolated in independent bags. What becomes of the FS number? Organizations might continue to use the original FS number on all of the bags, use the original FS number with a suffix added, or create entirely new numbers to track these new containers. During analysis, the contents of bags may be further modified. For example, an analyst may subdivide a bag into finer categories or isolate additional special finds that are deemed significant for analysis.
In most cases, they probably do not. Archaeologists assign systems of notation that are useful for analysis and later surrender the objects for curation without an understanding of how those numbers might be translated into accession or catalog numbers. A museum professional must process the collection and either adapt these numbers to the museum's catalog or generate new numbers by some other rationale. Depending on the institution, an archaeological collection's catalog numbers could hypothetically represent any type of division within the collection.
For example, a registrar might receive a site collection for curation. The collection has been thoroughly sorted by multiple archaeological specialists and analysts. It includes 500 individual bags of lithics (sorted by stone type) from 250 different proveniences, 100 bags of ceramics (sorted by temper type) from 50 different proveniences, and 1000 bags of animal bone (sorted by species) from 250 different proveniences. Field records and other files are surrendered for archiving at the same time.
How should the registrar process this collection? By what criteria should the registrar number and process these lots of specimens? What format will the FS numbers (assuming they exist) on those bags take in the museum catalog? Should be they given sequential numbers, numbers with suffixes, two or three-number designators? What systems of provenience were used at the site and what forms of notation do they take? Was their provenience information copied to the new bags when the collection was sorted? How will the link to the field records be maintained? Will it be feasible later to reconstruct what lithics/ceramics/bones were found together from any specific provenience? How likely is it that when the registrar finishes, that this collection will be easy for a researcher to query, access, and retrieve specimens from?
None of these questions will be easy for the registrar to answer and the solutions imposed by the museum professional may not be ones that archaeologists find desirable. A brief survey of a few publically available processing manuals of archaeological collections is provided here. The survey illustrates that there are effectively no standards as to how specimens are assigned accession/catalog numbers, what metadata is included, and some do not provide any sense of what exactly these accession/catalog numbers might mean. For example, an accession/catalog number could reference the entire contents of a site collection or could refer to some subdivision of the collection (e.g. a group of objects that share a common provenience, common material type, or other shared characteristic.) As a result, a site collection might have a single number for the entire collection or many different numbers for individual features or levels of features. This is likely to depend upon the size of the collection and the complexity of the collection's organizational scheme.