Solving Inventory Problems with Archival Research

One of the most important tasks in a museum is keeping track of items in the collection. To properly care for a collection, you must know what you have and where to find it. However, staff in any museum—especially those with collections acquired over more than a hundred years—know that mistakes can occur and we can lose track of things. Museum staff often encounter lost or misplaced objects, items found in storage without catalog numbers or documentation, and confusing or missing information in catalog records. Today we have technology and record-keeping systems that help keep these types of errors to a minimum but many unsolved inventory issues from the past remain at the National Museum of the American Indian (NMAI). 

Before computer databases and regular photography of all collection items, it could be difficult to determine when an item was truly missing or to match an unnumbered item found in storage to an existing catalog record. At the Museum of the American Indian, our predecessor institution, catalog card descriptions of objects were often vague. The cards usually only identified a culture and basic object type and did not physically describe the object, which often make it practically impossible to rectify inventory discrepancies. Without computerized records and accurate inventories, staff had to be certain of what was missing and also hope that a painstaking search through paper catalog cards might reveal a match. When staff couldn’t find a match and gave up the search, they would catalogue the item with a new number and note that it was “found in the collection.” That was the best they could do to ensure the object would be documented and tracked in the future. 

Typical Museum of the American Indian, Heye Foundation catalog card

At NMAI we are still dealing with longstanding inventory discrepancies. Our staff has made tremendous progress in sorting things out over the years but sometimes there are no clues or tangible leads to figure out an object’s origin or original catalog number. Fortunately, the Retro-Accession Lot Project—NMAI’s provenance research project to reconstruct the acquisition history of the museum’s collections—has provided new avenues to explore. Digging into the archives in search of provenance information has unexpectedly revealed solutions to longstanding inventory problems, including some we thought could never be solved. 

Haida figure from British Columbia, 24/8864. National Museum of the American Indian. 

In 1974, an unnumbered Haida figure was found in MAI storage: it could not be matched with records of any missing objects so it was assigned a new catalog number. Its new catalog card read “Carved wood figure of a kneeling man. Possibly a canoe prow or feast dish base, from collection - original number lost.”  

During the Retro-lot project review of an early scrapbook of clippings about American Indians and artifacts made by MAI founder George Heye, there was a familiar image: an article titled The Man Otter included a photograph of the figure found in the collection in 1974!  

Excerpt from Heye Scrapbook of Newspaper Articles. Museum of the American Indian, Heye Foundation records, Oversize Box 2. NMAI Archives Center. 

Using this information, a collections database search retrieved a catalog record for an object described as an otter man totem, possibly Haida. The object had been purchased by George Heye in 1904 from Frederick Landsberg, owner of Victoria, British Columbia, art and artifact shop called Landsberg’s Free Museum, and its original catalog card even referenced a newspaper article. This “totem,” which had never been photographed by the museum, had been considered missing for many years but had clearly been in the collection all along; it was simply obscured by its new number and catalog card that no longer referenced the “man otter.” 

Top: Catalog card for Haida Figure 1125 (now 24/8864). Bottom: Catalog card made during re-cataloging in 1974

Without additional documentation, it would have been impossible to determine that these two catalog records were for the same object. And if not for this chance discovery, made while flipping through a scrapbook in the NMAI Archive Center, we may never have solved this inventory issue.  

A:shiwi (Zuni) jar from New Mexico, RP0104. National Museum of the American Indian.

This A:shiwi (Zuni) jar from New Mexico provides another example: it was also found in collections storage with no catalog number; it was assigned a temporary number and housed with other inventory problems for years. Archival documentation regarding the 1906 acquisition of the Lewis Hotchkiss Brittin collection finally revealed its background. Brittin, a noted collector of books and Native American art, maintained catalog cards of his collection and often pasted photos on the cards. When Brittin sold his collection to George Heye, the cards were included, but unfortunately they did not include Brittin’s name or any reference to the MAI collections they might represent.  In 2019, we identified the cards as the collection purchased from Brittin in 1906 and matched them to catalogued objects. One card for a Zuni jar could not be matched to any catalogued items, but inventory records showed that a Zuni jar from Brittin was listed as missing. A search through the inventory problems revealed the missing jar, whose distinctive shape and design made its identification obvious.  

L.H. Brittin Catalog Cards and Plates. Museum of the American Indian, Heye Foundation records, Oversize Box CC3. NMAI Archives Center. 

Finally, this poncho from Peru has been a cataloguing mystery for years. It is a significant item because it shows evidence of the relationship between two Andean cultures from different regions of Peru, the coastal Moche and the highland Huari; its weaving and stylistic techniques represent both cultures and date it to AD 700–900. The poncho has been exhibited several times and been treated by museum conservation staff several times. Given its beauty and significance, it would have been a notable addition to the collection whenever it was acquired but it has never been clear exactly how or even when it was acquired. This is its long and convoluted story.  

Huari (Wari) poncho, AD 700-900, 24/4999. National Museum of the American Indian.

In 1976, the poncho was found stored in the museum vault and examined by MAI staff, who found that it lacked a catalog number. Due to its poor condition, it was recommended for conservation treatment in 1978 and was treated again in 1985. Documentation from this period refers to the object only as an unnumbered Moche-Huari poncho. In 1987, MAI staff concluded that the poncho had never been catalogued and assigned it a number: 24/4999. Given the poncho’s remarkable nature, some staff later felt that it must have been catalogued earlier and lost its number. Over the course of decades, they investigated missing archaeological textiles from Peru to try to identify the poncho’s background but met with no success. 

What complicated later efforts was that the catalog number 24/4999 suggested that the poncho’s acquisition dated to 1971. MAI (and NMAI) catalog numbers are sequential—1 to 270600 currently—and rough acquisition dates can often be ascertained by their placement in that sequence. The poncho’s catalog number fits in the range of numbers assigned in 1971 and, from the 1990s onward, staff assumed it was acquired before 1971. What they did not realize was that in the 1980s, staff had identified a block of unassigned numbers and used them for newly catalogued items, including the poncho. In the early 2000s, when staff attempted to match the poncho to a missing object, they only scrutinized objects catalogued before 1971. 

In 2019, the origins of the poncho finally became clear when it was matched with records for a 1973 purchase of what was described as “an extremely rare Mochica poncho from Peru.” The purchase—arranged by director Frederick J. Dockstader from Swiss antiquities collector Jean Lions—was significant enough to merit mention in the museum’s 1973 Annual Report. MAI supporter Harry Blumenthal had provided funds for the poncho’s purchase in the name of Arthur Sackler, a generous donor who had started an MAI fund to support purchase of unique and outstanding objects. In 1974, Dockstader catalogued the poncho with the number 24/8860, but that number was never written on or attached to the poncho itself, leading to decades of confusion.  

Original catalog card for Huari (Wari) poncho 24/8860 (now 24/4999).

In 1975, MAI dismissed Frederick Dockstader following a New York State Attorney General’s Office investigation of sales and irregular deaccessions of MAI collection objects. In 1976, when the poncho was found in the vault, no one sought an explanation from Dockstader, and during the inventory of the MAI collection in the late 1970s, poncho 24/8860 was never located and was assumed to be missing. 

As part of NMAI’s Retro-Accession lot project, we have not only worked to reconstruct how collection objects were acquired but also to understand MAI’s history as a museum and its practices. We can never take for granted that the standard museum practices we follow today actually pertain to earlier decades and understanding how MAI operated has helped us unravel some of these more complex situations. Looking back at the chain of events, it’s now clear how the poncho lost connection to its documentation and how staff repeatedly went astray in earlier efforts to determine its acquisition history. The poncho’s story also demonstrates how easily we can lose track of items and how changes in staff and loss of institutional knowledge can contribute to what we know about items in the collection, all of which shows that good and timely record-keeping is critical.  

Armed with a better understanding of our museum archives and collections documentation, we now have new tools to solve old problems, including persistent inventory issues and ensuring that collections data is accurate. For more information on the Retro-Accession Lot project and how it is changing what we know about the NMAI collections see previous blog posts here and here.  

Maria Galban 

Collections Documentation Manager 

National Museum of the American Indian