SNAC Records Workflow

Owned by Betts Coup
Last updated: Jul 02, 2024
8 min read

Houghton Library SNAC Workflows

Prepared by Betts Coup & Melanie Wisner, 2019-2020, updated 2022 and 2024.

Introduction:

Social Networks and Archival Context (SNAC) is a platform for describing contextual information about creators of archival materials. SNAC brings physically disparate materials by the same creator together intellectually. Houghton Library is thus far the largest contributor to SNAC and is committed to creating new records as well as maintaining and cleaning up legacy records in an effort to enhance discovery of archival materials. 

The Houghton Library SNAC Policy outlines three levels of detail for describing creators and contributing records to SNAC. These levels are in accordance with established standards for authority records as well as SNAC’s own requirements for validation.  

Workflows below outline how and when to create each type of record for the first time for each of the three levels, followed by procedures for how and when to edit and merge existing records.

All SNAC records created by Houghton staff should be listed in the Tracking Spreadsheet. For other helpful tips and reminders for doing SNAC work, please see this document.

To see screenshots within the SNAC workflow, please visit this document.


Three Levels for new SNAC Constellations:

Level 1: 

Accession level

  • Core data: names (preferred and unpreferred), with date range if available
  • Biographical note 
  • Exist dates (if available)
  • Resource relation
  • Sources
  • No clean-up of any fields.

Level 2:

Note: Records must be level 2 or above to be eligible for merging.

  • The above, plus:
  • Places associated with creator and records
  • Occupations
  • Subjects
  • External CPF (Same As) , usually a Wikidata image, if available
  • Clean-up of the above fields, though clean-up of Sources is optional.


Level 3:

    • The above, plus:
    • CPF relations

Workflow for SNAC records:


At accessioning, the aim is to create a (or edit an existing) Level 1 SNAC record for the main creator and any other significant figures associated with the archival collection if not already present in SNAC. If a SNAC record already exists, accessioning archivist confirms accuracy, adds resource link, and, if necessary, flags entity for Level 2 or 3 treatment. 

When processing, SNAC Levels 1, 2, or 3 may be used; these do not necessarily match up with levels of processing. The archivist should consider the creators’/collectors’ scope, historical stature or lack thereof, depth of meaning to Houghton collections, and other criteria when deciding on the level of SNAC record. The vast majority of creators will receive Level 1 records, without clean-up.

Detailed instructions
Creating Level 1

Ensure that the entity does not exist in the system via the browse function. Check both surname and forename-first entries, just to be sure. 

If the entity does not exist, go to the Editor’s Dashboard and click Create New Constellation.

  1. Create core data (Name entries tab): 

Core data in SNAC is where you will list the preferred and nonpreferred names (with life dates) for the entity you’re describing. All names should include life dates if available, and SNAC prefers the fullest form of name, even if this is not the LCNAF authorized name for the entity. Adding variants is not required, but is recommended if an entity regularly used a different name or their name regularly appeared in another language or script.

  1. Entity type: select person, family, or corporateBody
  2. Name entries (SNAC adds punctuation)
  3. The preferred name should always have the “authorizedForm” code, even if this is different from the LCNAF authority.
    1. Parse the name into Surname, Forename
    2. Add dates
    3. Set preference to preferred
    4. Assuming language is English and script is Latin, hit the magic wand. This will auto populate these fields. If the language and script are different, add these manually.
    5. Set rules to “authorizedForm” and “rda"
  1. Add name expansion if necessary
  2. If necessary, make additional name entries for variant names
    1. If you have dates, always include them in every variant name
    2. All variants, even if it the form in the LCNAF authority, should be coded with “alternativeForm” and “rda.”
  3. Once you have added all names, click continue. It will give you a list of possible entities to make sure you haven’t created a duplicate. If you have, select edit, and you can edit that entry and the new entity will not be created. If there is no duplicate, scroll to the bottom and click Create New Constellation.
  4. If you are editing a record, you can but are not required to update all name variants: delete duplicates, add life dates, and add coding

2. History tab

The History tab includes data about the biographical history of the entity you are describing, including a biographical note, life or establishment dates, and places to which the entity had a relationship (birth, residence, death are common types of relationships). 

  1. Bioghist
    • Copy and paste biographical note from finding aid or MARC record, or if there is not one available there but there is one on Wikipedia, copy and paste from there. Include life dates if available
    • If you’re editing and there is a bio note present, read through and clean it up. Remove any citation content and move that to sources!
    •  Code for language and script as above
  2. Exist dates
    1. Use date range for exist dates if available
  3. You must fill out the Machine-Readable date fields and the date type. You can add natural language dates but that’s not necessary for the system
  4. Use a single date if that’s all that is available

3.  Relations tab

The Relations tab includes the three types of relationships you can fill out in SNAC - relationships to archival and other types of materials (Resource Relations), other entities (Internal CPF Relations), or other authorities for the entity (External Related CPF). Houghton requires the addition of a resource relationship when working on any SNAC record; level two records may include external related CPF relationships; level three records will include internal CPF relations. 

  1.  Click to add a new resource.
    1. Generally our Level 1 resource relations will be creator or collector relationships, but a person or corporate body serving as the subject of a collection would be coded “referencedIn”
    1.  Ensure resource does not already exist in the system
    2. Add required elements: resource type, display entry, and title
    3.  If available, add date, permalink for HAD resource, abstract, extent, and languages – copy from finding aid or MARC record (link must be copied from MARC record)
    4. Click continue.
    5. Select role: creatorOf, contriborOf, editorOf, referencedIn
    6. Add repository:  Houghton Library 

4.Sources tab

The sources tab is where one records all the sources used specifically to create the contents of the SNAC record. These will commonly include finding aids, Wikipedia or Find-A-Grave websites, or other web entries, but may include bibliographic sources. Sources should be added only if they provided information that has been used in the SNAC record, and that information will be included in the Found Data field for each source. This includes finding aids - all collections should be included as a Resource Relationships, but a finding aid should only be listed as a source if its content was used for the SNAC record itself.

  1. Sources
    1. Source name (short title of the source)
    2. URI (optional)
    3. Source citation field:
      1. Formatted in Chicago style for book citations
      1. Use finding aid title to cite a finding aid
      2. Use the preferred citation if citing actual collection material
      3. Wikipedia and Find-a-Grave citations can look more like this: “George Read, Wikipedia entry, accessed December 2, 2021”
      1. Full bibliographic citation for the source
      2. Can be more casual for online citations
    1. Fill in the following sections:
  2. Found Data field:
    1. Detailed information about what was found in the source to support specific assertions
    2. If one source, it’s recommended to copy and paste all content from that source into Found Data (biographical note from finding aid/bib record, Wikipedia biography).
    3. If there are multiple sources, include what particular information was pulled from each source. Recommended formatting is phrases separated by semi-colons, mimicking an LCNAF authority.
    4. If copying the content of a biographical note with considerable XML coding, the Found Data field will not recognize it. Ignoring <p> tags is fine, but remove longer lingering strings of XML when copying into Found Data.
  3.  Language and script
  4. Always include our finding aid and possibly a Wikipedia entry
  5. Clean-up of sources is not required. You can delete sources that are just pieces of XML or don't point to a source, but this is not required at any level.

Creating Level 2 

  1. All of the above, plus:
  2. History tab
    1. Add new place
    2. Select the appropriate type (AssociatedPlace)
    3. Select the appropriate role
    4. Add a descriptive note if necessary
    5. If the desired place does not already exist in the system, contact Jerry Simmons
    6. Places
  3. Demographics tab
    1. Select appropriate occupations from menu
    2. Add if necessary
    3. Use occupations listed in MARC record or finding aid
    1. Select appropriate subjects from drop-down menu
    2. Use subjects listed in finding aid or MARC record but only those that apply best to the entity, not the collection
    3. Occupations
    4. Subjects
    5. If editing, check the other options, especially Languages Used and make sure they are filled out properly, no need to add. We do not use the Genders and Nationalities options for SNAC at Houghton.
  4. Relations tab
    1. Extermal CPF (Same As) to add for image from Wikidata or to link to an external authority file (LCNAF)
      1. Click the +Add New Same-As Relation button and use the entry screens to link to a matching entry in an external authority source
      2. Select the type of same-as relationship. Typically this will be Wikidata or LoCNAF.
    2. Enter the identifier.
      1. For Wikidata item, go to Wikipedia page for the entity and then scroll on the left until you see the Wikidata item link; click and go to that page; copy for the specific Wikidata item number.
      2. For LocNAF, copy just the identifier (N plus several numbers) at the end of the URI
    3. The URI will automatically be filled in.

Creating Level 3
  1. All of the above, plus: 
  2. CPF relations - Relations tab
    1. Choose internal CPF relations: 
    2. Add new Constellation Relation
    3. Choose constellation that you want to connect (SNAC constellation must already exist)
  3. If above constellation does not exist, create at least a level one record and then create relationship
  4. Select relationship type
  5. Add descriptive note if necessary
  6. Add the relationship to this entity in the other entity’s record, also as an Internal CPF Relation

Workflow for editing existing SNAC records

  1. Go through all the above steps for the level of record you are working on, editing rather than creating anew
  2. For Level One records: do not clean up! Just add our content as listed in the instructions above.
  3. If Level Two or Three:
    1. If a biographical note is already present but includes citations, make sure to remove the citation from that field and add it as a source. You may need to search for a finding aid on another repository’s site for this purpose.
    2. Check resource relations and ensure they are up to date
    3. Update Houghton resource relations
    4. Do an efficient assessment of existing data for accuracy 
      1. One of the issues of SNAC is that multiple creators can be conflated and described in a single constellation. Remove any data that is about a different entity
      2. Assess biographical notes for biased or outdated content
    5. About WorldCat identities: editors may replace WorldCat Identities links with links to our finding aids
    6. Delete excess invalid sources (XML coding) as noted above

Workflow for merging SNAC records

  1. When you are updating existing SNAC records and notice that there are multiple constellations for the entity you are describing, merging is necessary. Merging should only be done in cases where you are certain that the multiple constellations are describing the same entity
    1. If a heading appears vague, follow the  links to Archival Collections, Related Resources and Related Names in SNAC to confirm a match. You may have to scan linked archival finding aids and CPF links to finalize your merge decision
    2. If you are uncertain, you can just highlight both and click Add Maybe Same, which will tag both records as likely the same, but will not actually merge
  2. The first step of merging is the actual merge of the records
    1. In the Browse function, where you can see repeated names, designate the records to merge by checking the box to the left of the names
    2. Once you’ve decided on the records to merge, click the Auto Merge button at the top right
    3. Note: you don’t have to merge them all at once. You can choose just two or three at a time as you pare down the list of duplicates
  3. The final step is editing the newly merged record - Level One records do not require editing.
    1. Please note that ALL of the headings from ALL of the merged records are now in the new main record
    2. Edit away, or trash any duplicate name entries from the Core Data tab of the new record
    3. Check other tabs (History, Demographics, etc.) for other duplicate entries