Vol. 9 No. 8 — September 2001
The Science and Art of Online Research in the Fine Arts:
A Process Approach
by Mary Colette Wallace • President, The Wallace Research Group
Table of Contents Previous Issues Subscribe Now! ITI Home
Process involves making sense of what is known so that the unknown can be explored. The confusion of online art research arises from the multiplicity of subjects such as ownership, location, category of artwork, history, and other aspects involved in pinpointing a certain artist or artwork. When one chooses to avoid fee databases such as Dialog, with its formal searching structures, the searching tasks multiply. The process generally covered here is just one of many methods that enable a coherent organized search. 
"Research is the process of going up alleys to see if they are blind."
—Marston Bates
Herbert A. Simon wrote in The Sciences of the Artificial (Third Edition, The MIT Press, Cambridge, Massachusetts, 1996, P. 211), "There is now a growing body of evidence that the activity called human problem solving is basically a form of means-ends analysis that aims at discovering a process description of the path that leads to a desired goal. The general paradigm is: Given a blueprint, to find the corresponding recipe." 

Searches in the field of art research may involve multiple online databases, so trying to program a search according to the fields each database assigns requires too much peripheral information and time. Upon reflection the analogous question becomes, "Is it better to spend time preparing and improving swim techniques — OR — to spend time on memorizing the pool dimensions, depths, and hours of operation for all the different swimming pools I might use?" 

The optimum process requires understanding the function of the needed and given information before taking action. Howard S. Becker, in his wonderful book on the process of research, Tricks of the Trade: How to Think About Your Research While You're Doing It (University of Chicago Press, 1998, p. 44), writes, "Making activities the starting point focuses analysis on the situation the activity occurs in, and on all the connections what you are studying has with all the other things around it, with its context." Such connections for art researchers are precisely essential to the better terminology, more accurate querying, and higher-quality retrievals.

The "science" of online art research lies in organizing the facts about an artwork, by asking a few simple, easily remembered "connection questions" first from the client's standpoint and then from the resource standpoint. 

  • Who is/was the artist, or client or subject?
  • What is/was the medium, technique used?
  • When in time does/did the artist live, create the artwork, die?
  • Where does /did the artist live while creating this artwork?
  • Why is/was this artwork created?
  • How much is it worth, as in asking price, value, actual sales price?
Once these questions are answered, the next step is to prioritize them according to the clients' request. In all likelihood simple searches won't require all six questions answered or verified, but the entire process of asking them and considering the answers to the questions remains an invaluable thinking guide that automatically yields key concepts and primary search terms to prioritize.

Apply the same connection questions to the source side of the equation. Programming the search will help to identify the type of information needed, but the next step is to identify the appropriate sources. Optimal sources are those that gather the information, maintain it for their own needs primarily, and then secondarily allow others access to it for free or for a fee. By asking the connection questions you will have a better view of the possible sources for the information you're seeking.
Typical Client "Connection Questions"
Who (image search, museum/collection search, auction /sales search, provenance search)

"I have an image and I don't know who painted it." 

"I need to find out whose portrait this is."

"I need to find out who owns the copyright on this painting."

What (museum/collection search, auction/sales search)

"I know the name of the artwork, the artist, and where it is now, but I need to know if it was painted on vellum, paper, wood, canvas, or glass, etc." 

"I need to find out if it was ever included as a plate or illustration in a book." 

"I need to find Baroque paintings of flowers."

When (museum/collection search, auction/sales search)

"I need to find when this painting was painted (period, style, actual date)."

"I need to find out if this artwork was painted during the artist's "blue period.'"

Where (auction/sales search, exhibition catalogue/catalogue raisonnes, provenance search) 

"I need to find the current location of a certain painting." 

"I need to find out who owns the copyright." (First it will be necessary to find out where the artwork resides.) 

"I need to compile a list of all the owners of an artwork (provenance)." 

Why (museum/collection search, exhibition catalogue/catalogue raisonnes)

"I need to understand the choice of medium for these specific years (during which there happened to be a war that limited certain supplies or perhaps the artist acquired a wealthy patron and improved the quality of their art supplies, etc.) 

"I need to find out if Vincent van Gogh made a living by selling his paintings."

How (auction/sales search)

"I need to find out how much this painting sold for 20 years ago."

"I need to find the current value of a painting."

Source "Connection Questions"
Who = Who collects and uses information of this type? Private and public individuals and entities such as collectors, appraisers, dealers, auction houses, museums, companies, libraries, universities, designers, architects, companies, investment and tax consultants, researchers, database aggregators, publishers, and so on.

What = What information do these entities make accessible — for free or for fee — to the public? In what forms — CD-ROM, print, microfiche, online, or a combination of several? What subjects do these entities specialize in?

When = From what date to what dates do the records in this reference source cover? When and how often is the source updated?

Where = From where do these entities get their information? Has this information been a part of another database or reference source in the past?

Why = Why do these entities obtain the information? Do they use it for education, display, investment, collection, administration, or just for a tax shelter?

How = Does the source buy, donate, or sell art? With public and/or private money? 

Art Research Source Information Flow
Art Research Source Information Flow
The Art research process, graphically represented.
The Art research process, graphically represented.

Depending upon search priorities, a starting point might require different databases — a museum collection rather than an image library, a provenance database, a timeline filled in with each painting sold at auction, or all source types in various combinations. What you retrieve may require a reorganization of priorities, given the high fluidity of online art resources. 

"A line is a dot that went for a walk." —Paul Klee
Browsing is as essential to the art search process as "serious" searching. Process teaches that fact-finding questions will more likely advance solutions, but also can enable artistic intuitive insights or leaps. Browsing presents a broad view of the search subject, a mere dipping of toes into the pool to check the temperature of the water. A quick search on the name of the artist or the subject (if a portrait) provides broad instantaneous clues such as level of coverage by chosen search engines; major museum, gallery or library collections and arts directory sites to investigate; languages other than English involved; and period or style. Receiving few retrievals or no retrievals from an artist's name search doesn't necessarily indicate the quantity of information available, but it may well indicate the advisability of expanding target terms to more interdisciplinary, multi-lingual thesauri.

The comparison of retrievals from browsing improves the mental sifting needed to evaluate the most likely sources according to the connection questions. This activity is the mental equivalent of "priming the pump" — getting enough pressure to raise water from the deepest well. A scan for interesting resources in the retrieval citations might provide the client with the information they need and leave nothing more than the verification to complete. But for more complex or illusive searches, this is a good time to get familiar with the contextual settings of the artist and his or her work, along with the type and quality of information likely to be found. 

At this point the search is like dipping your toes in the water of a broad reflecting pool that's only one-half-inch deep. The "facts" provided by the client may well be accurate — but you need to verify them. Caution, reflection, and attention to details are all good habits, even when verifying the target information at Web sites of the most reputable museum, library, or university collections. Verify the spelling of names, geographic locations, titles of paintings, and dates, whenever possible. Additional languages and alternative spellings of artists' names and even of everyday words, such as watercolor (American English) and watercolour (British English), will provide more target term options. Watch for any discrepancies in spelling, in dates, in any of the information the client gave you. If you find any such discrepancies, note them and re-check sources before going on to other reputable sites. Finding any differences doesn't necessarily indicate fraud, but the field of art research always involves concerns about authenticity.

In a recent search, I found that some databases/Web sites/collections place the Dutch 17th century as part of the Baroque period, while others consider it part of the Renaissance, making those terms more troublesome. So, while the term "17th" will enable me to narrow within a search, I will not use the terms "Renaissance" or "Baroque" except within certain directories and indexes which require such a choice to drill down for the artists' name. At that point, it's easiest to choose one, e.g., Baroque, take a quick look at the dates, and, if necessary, go to the other. As to searching exact dates, since the artist in my example — Pieter Withoos — is rather obscure and the client requested information doesn't require it, I won't use the birth-death dates — 1654-1693 — except for verification.

"Art washes from the soul the dust of everyday life." 
—Pablo Picasso
There are so many new "facts" replacing old "facts" with the aid of technology and interdisciplinary curiosity by researchers, that time considering disputed or intuitive connections is well spent. Verified facts form the essential structure against which to compare conflicting facts or information. For example, the famous painting the "Mona Lisa" by Leonardo da Vinci — also known as Leonardo de ser Piero da Vinci — is also known as "La Gioconda," "La Joconde," and "Mona Lisa del Gioconda." These usually country-specific spellings are useful alternatives for searching in databases not returning any hits on anglicized spellings. A search on the "Mona Lisa" misses out on some great European art sites and loads many unwanted retrievals (although I do love Nat King Cole, too). 

Unverified facts can lead to confusion when there are different language versions of the same item, such as the title of a painting by the 17th century Dutch artist, Alida Withoos, "A Forest Floor Still Life with Various Flowers." The same painting image can also carry the title, "Blumen in Einem Waldgrund." Given that the online image is obviously the same, and even with a rudimentary knowledge of German, one can easily verify a translation has occurred. But which one is the original? Is one a translation provided at an auction? A verification or clarification might be in order.

"The most beautiful thing we can experience is the mysterious."
—Albert Einstein
Terminology about a specific item can shift from industry to industry. In fields other than art, a drawing and/or an inked watercolor painting might be regarded as an illustration or simply a drawing. Then again, there are historical usages of terms such that a watercolor and ink painting of today might once have been termed a drawing as a result of the type of method and/or the purpose of the artwork. A pen and ink study or charcoal cartoon of the 17th century might be regarded today simply as a drawing or even as a painting. Aspects of terms such as these can uncover hidden sources. However, an over-reliance on specific terms can also lead to information — though sometimes fascinating — that does not serve the client's needs. 

Multilingual terminology can encompass varying nationalities, disciplines, and alternative descriptions and provide additional depth to primary search terms. The selection of languages as well as the terminology is part of structuring the query that increasingly requires more time. Partly driven by the use of two or more languages and partly by interdisciplinary searching, terms can make or break a research project with too much complexity or not enough. For help, try a good thesaurus directory such as the subject-arranged Web Thesaurus Compendium [

Upon finding several of Pieter Withoos' artwork titles in French, German, and English, I decided to try a key term in those and Dutch language versions of search engines. Using the German version, I retrieved a stunning new auction site in English ­ the Gabrius Auction databank [] — with some images hidden from any English-language-only Web search engine.

"Imagination will often carry us to worlds that never were. But without it, we go nowhere." 
—Carl Sagan, Cosmos
Graphic communication increases the numbers of multi-lingual viewers for Internet sites. Responding to this growth, many search engines such as Ditto, Google, AltaVista, and Lycos are racing to improve searchable access to online images. Photo aggregators and resellers of images, such as Corbis, have site-specific, easily searchable image databases. But the odds aren't good for the basic Web search engines if they have to handle approximately 1,346,966,000 searchable Web pages, like Google does, and if each page has at least one graphic such as an arrow or a "back" button. Current keyword searching for images in Google results in nearly half that are not artwork images, but "gif" images — Web page graphics such as bullets, back buttons, and the occasional logo given a text tag for greater accessibility. Part of the difficulty lies in using text to find images. Not only a language barrier, but more importantly, a cultural barrier exists in deciding what an image primarily relates and how that can be summed up in a keyword. 

Visual art — graphic and fine — is one of the purest forms of communication in that it transcends verbal language barriers. Many times, it forms the only and the deepest connection with the artists themselves. The artists' interpretations hold intensely powerful and subtle views of history, styles, music, architecture, wars, politics, love, pressures, joys, and beliefs of their time. Viewing the image provides a lot of visual clues for alternative keywords as well as the "scent of the hunt" for those inevitable moments when you have to struggle to find anything in database after database. 

In many image searches, the artist's last name is the primary term to search on or find via an index, e.g., "Withoos." The importance of the artist's name in any image search dictates that several configurations be tried: the last name followed by a comma and the first name, e.g., "Withoos, Pieter" or even "Withoos, P.," as well as the modern first name then last name order, e.g., "Pieter Withoos." If you can't locate the specific artwork image, then look at similar works by the same artist, on the same subject if possible (e.g., landscapes, portraits, etc.), and in the same medium — to get a sense of the artist's style. An outside possibility exists that the first name is misspelled, e.g., "Peter" not "Pieter," and that approach too might be worth a quick search.

"Art teaches nothing, except the significance of life." 
—Henry Miller
The more obscure the artist or work you are searching, the more important it becomes to select carefully good, interdisciplinary terms and to prioritize them. Several museum and collection sites have searchable fields that include artwork type. Outside of such scholarly sources, the use of terms such as "painting" or "drawing" is usually far too general, with different meanings within different sources. The nouns "drawing" and "painting" will bring too many contemporary and other unusable retrievals as they are also verb forms ("He is drawing conclusions," "She is painting the town red"). 

Strictly as an art query, your best approach is to focus upon the more unusual and prioritize accordingly: name (if it's unusual as in the example "Withoos" rather than "Smith"), then century (17th), then nationality (Dutch), then type of artwork (study) or movement (such as Impressionism). In the example, the artwork title of study indicates the picture might have been included in a book or mentioned in botanical works, so a comprehensive search would also cover botany and botany-related databases, as well as art-related databases.

Searching on the specific artwork title subject, e.g., "study" and "gourds" or "drawing," "painting," "watercolour," "watercolor," will help retrieve a list of artwork citations. The interdisciplinary search terms in the example are botanical and botany-related such as "illustration" (as opposed to the art term "drawing" or "painting") and "kalebas," "mergpompoenen," "cucurbit," and "gourd." Using botanical terms, I find that Pieter Withoos is the "author" of a book that consists of plant drawings and paintings. Due to the artist's life dates, it is likely that the book production involved hand tinting, which would greatly reduce the number of printed copies and any online access to individual prints or plates. 

"No amount of skillful invention can replace the essential element of imagination." 
—Edward Hopper
Due to the constantly changing nature of the Web, there can be no guarantee of online access to sources — even to those with subscriptions, and perhaps this is one of the most profound differences in online accessibility versus printed matter accessibility. Once a book is bought, it remains intact and accessible until it's physically removed or destroyed; once a database subscription is signed for, there are no guarantees that access will be there tomorrow or even the next hour. Yes, most important content is sold to another company, but subscribers don't always receive notification. 

Ironic, but an absolute must-have resource for doing online art searches is a book by Lois Swan Jones, Art Information and the Internet: How to Find It, How to Use It (Oryx Press, 1999, Phoenix, Arizona). It is an excellent, highly organized resource, and the publisher provides a Web site where you can find updates for the untold numbers of URLs in her extensive book. 

Part of the difficulty of art research stems from the awesome quantity and quality of art sources online from art directories and indexes to museum collections and provenance databases. It's such a delight to virtually "visit the Louvre" or the "Met," to "see a world in a grain of sand and a heaven in a wild flower," as William Blake wrote in his "Auguries of Innocence." Priceless too are the location histories of each work taken by the Nazis that are now finding their way back to the owners through searchable provenance and spoliation databases. 

Just a word on the giant subjects of provenance, provenience, and spoliation sources. Provenance (or provenience — a rarely used derivative) is the result of research to prove the past right and legal ownership and locations of an artwork with the goal of establishing the value and authenticity of an artwork or of restoring the artwork to the authentic owner and/or their heirs. The latter form of provenance research is mandated by law in many nations in efforts to rectify the billions of dollars' worth of artwork looted from Jews by the Nazis in World War II, which was then illegally "gifted" or sold to museums, collections, galleries, universities, libraries, religious institutions, and individuals around the world. Spoliation, meaning "the spoils of war," is a term used primarily in European art museum, collection, or research databases, as an alternative to the terms "provenance" or "provenience." 

Major museums such as the Getty, the Met, the Frick, and so on have searchable provenance databases at their Web sites. UNESCO has a project underway to post international laws "governing the control of cultural property" for museums, collectors, dealers, and customs officials at the Web site [], as well as international sources of information and links to databases. The United States National Archives [] has information on provenance research and principles with special coverage for the Holocaust era. 

Finding artworks can put a surprising strain on a searcher's inventiveness. Licensed artwork sources can extend from botanical books to stamps to dinner plates, as well as traditional sources such as museums, galleries, and collections in universities. One good way to assess critically the veracity of an online art source is to consider what the source would risk if their information on an artwork were incorrect. 

Before the final searches commence, the list of interdisciplinary sources needs to be prioritized with the idea of going from the broadest to the narrowest. Due to a widely disparate overlapping of online searchable fields and collection depths, the more optimum general approach is to focus on the auction — sales catalog sites, provenance sites, and image galleries. If the preliminary search retrieves citations for major museums, galleries, or specialized collections, these would qualify for the next tier of databases to search.

"We live in a beautiful and orderly world, not in a chaos without norms, even though this is how it sometimes appears." 
—M. C. Escher
The filtering of search retrievals may seem a very straightforward answer to one of the connection questions such as, "Who owns this painting now?" If the owner has been found, then the verification process is a quick finalization. For those still unanswered questions, the process can be thought through again. Only a miniscule percentage of the combined total artwork residing around the world in libraries, universities, museums, galleries, and other collections is available online in the collections of museums, libraries, and other major owners. So the end of the search online may be a terrific start to a search offline through an identified collection's curator or printed source.
"What good are computers? They can only give you answers."
—Pablo Picasso
Art research is as much a thinking and visual process as a textual process. Widely varying online resources require a process structure for searching multilingual interdisciplinary subjects where the most reliable information currently appears. While there is still a huge backlog of museum image and collection resources yet to be made available online at many art related sites, newer invaluable tools of provenance and art auction databases have greatly enhanced the online art search. 

Note: All the artist quotes used appear on

Art Directories, Indexes, and Image Databases
Art Links



The Artist Index


The Axis Database Online


Bubl Link: 700 the arts

Dutch Art Online

ECIT — Electronic Compendium of Images and Text — The Piero Project 

Finding Images Online

Internet Art Resources

Lycos Image Gallery

Voice of the Shuttle

The Warburg Institute

World Art Treasure

World Wide Arts Resources

World-Wide Web Virtual Library:  Art History



Art auctions and art as investment: see Files 9, 16, 148, 649, 47, 781 and the OneSearch category PAPERSMJ.

Art Index — The Wilson Art Index


The Art Sales Index database 2.4 million entries) 



Comision Para el Esclarecimiento de las Actividades del Nazismo en la Republica

The Bruno Kreisky Archives Foundation

Musees Nationaux Recuperation

The Prime Minister's Office

Schloss Collection, Non-Restituted Works Looted 1943-1998

Commission for Looted Art

Lost Art Internet Database

Interministerial Commission for Art Works

Vilnius International Forum on Holocaust-Era Looted Cultural Assets

The Netherlands
Museum-Security Organization

Swiss Embassy

United Kingdom
Cultural Property News

National Museum Directors' Conference

United States
Art Loss Register

Commission for Art Recovery

Getty Provenance Index

Holocaust-Era Assets: Research Institute Resources

Holocaust-Era Assets: Records and Research at the National Archives and Records Administration

Holocaust Art Restitution Project (HARP)

Los Angeles County Museum of Art

Metropolitan Museum of Art, New York

Museum of Fine Arts, Boston

Museum of Modern Art, New York

Project for the Documentation of Wartime Cultural Losses (The Documentation Project)

U.S. Holocaust Memorial Museum 
in conjunction with the Washington Conference on Holocaust-Era Assets 
International List of Current Activities Regarding Holocaust-Era Assets

U.S. State Department, Bureau of European Affairs Holocaust Issues

World War II Resources at the National Gallery of Art

Mary Colette Wallace is president of The Wallace Research Group []
Her e-mail address is
Table of Contents Previous Issues Subscribe Now! ITI Home
© 2001