The catalogue raisonné (CR), a comprehensive and detailed listing of all the known artworks made by an artist, is to the study of art what a dictionary and grammar are to the study of a language.

Unlike dictionaries and grammars, which can be compiled successfully by individuals or very small groups, the CR is often the result of many years devoted attention by many scholars. They not only have to document works which are generally accepted as being by that artist, but also have to act as the supreme arbiter of those which are genuine, and fakes. In doing that, decisions can make enormous differences to the value of paintings, which inevitably can become extremely delicate.

Some artists and their agents leave detailed documentation which is reliable evidence of date of creation, etc., of their works, but more often than not, those compiling a CR have to build a reasoned chronology of those works, and validate each step in the provenance through which the work comes to be where it now is.

The traditional end result is usually a series of hefty volumes, published by a specialist house (such as the Wildenstein Institute, a gallery, or academic art publisher) at very high cost. The more popular, such as Wildenstein’s CR of Monet’s oil paintings, may be offered more widely; Cézanne’s 1997 CR was published in three volumes by Thames and Hudson, and now costs £236. John Singer Sargent’s CR is due to be completed this summer, with the publication of the ninth and final volume by Yale University Press, at a price of £50 for that volume alone.

If you are interested in an artist, investing in a copy of their CR can be a huge step forward. Access to CRs in even quite large libraries can be difficult, and instant personal access lets you see things which the occasional rummage may not. But at those prices, few can afford more than a small number for those artists who are most important to them.

CRs are also prepared relatively infrequently, even for the most prominent and important artists. In many cases, including most of the painters which I have covered in the Peri-Impressionist and Vanished Impressionist series here, the most recent CR may be more than thirty years old, lack colour illustrations, and be almost unobtainable. In the case of more obscure artists, no CR may ever have existed.

Everything that I have written so far must drive towards one conclusion: that CRs need to go electronic. Some have: the most extensive and successful that I know about (and use) is Paul Cézanne’s, which is accessible free of charge to registered users. Other major artists whose CRs have gone or are going online include Mary Cassatt, Roy Lichtenstein, Tom Thomson, George Inness, Jan Brueghel, John Singer Sargent, Louise Bourgeois, and Edwin Dickson.

Surprisingly, the two major content management systems (CMS) which are most used for online CRs are commercial products. panOpticon starts at $5000, and its major competitors from GallerySystems are commercial CMS designed for institutional collections. It is puzzling that galleries have not got together to fund the development of an open source system, as has happened with Greenstone digital library software, for example.

Projects like Google Art and other online galleries are wonderful, but without online CRs as a basic building block, they are limited to the most popular works in major galleries. The information about works provided in Google Art and Wikimedia Commons is also sadly inaccurate when compared with recent scholarship: even such fundamental information as the medium and size of paintings is often completely wrong.

When I write articles about art history for this blog, one of the most time-consuming and frustrating tasks is compiling the citation for each image which I wish to include. On occasion, I have even discovered that an image attributed to a particular artist in Google Art or Wikimedia Commons is incorrect. Such problems must be even worse for those compiling works of real scholarship.

Online CRs are ready for an open initiative, to put our cultural heritage on the firm base which it needs. They are to that heritage what species genomes are to molecular biology.