english       hämta      

Sharing information and resources: cooperation of museums, archives and libraries in the field of standardization
National Library
Riga, Latvia, April 20, 2007

Detta webbdokument är publicerat och ingående länkar är kontrollerade 2007-04-30

Making one from four - and mastering the authorities

Some remarks on sharing information and resources from a Swedish museum perspective…


Different systems develop from different purposes and it is necessary to see the development of them as a result of a complex set of prerequisites. Outside the box there is not always a good possibility to reuse the tool in such a transparent way as it may seem at a first glance. There is a range of necessary questions to take in consideration.

We often think of technical issues here – such as machine platform, operating systems and the program environment of the databases - as the main obstacles in collaboration. This is not the whole truth, if even a problem at all.

We are today facing many more problems areas, which will in fact not be obvious to curators nor to technicians and systems architects which groups both tend to look at “their” favourite issues, of course.

Facing the increasing need for cooperative management of heritage knowledge – or at least possibilities to search and find information without being aware of the sources and their special structures – we need to look deeper into a field of meta-knowledge between the “real” knowledge and the machines and systems.

This is a field where heritage institutions haven’t been so focused in earlier days – the sources were local and the experts also. The same goes for many of the end-users. A large amount of understanding was built into the professionals routines and was transferred from person to person in order to be able to read old ledgers, cards and registers.

Standardising the databases force us to not only be experts on 18th century paintings, wooden furniture, old maps or a certain area of industry history. We also need to be able to handle the language in a standardised way, to be aware of syntax and structure, spelling and multilingual issues in a much higher degree than earlier.

I will here talk from a museum perspective, because this is my main field, but I think that most of the arguments and examples can be transferred to archives and libraries as well and I am convinced that you all can turn my experiences into your perspective.


Examples of different unstructured ways of writing information:
Hans Rengman, Uddevalla
Rengman, Hans, Uddevalla
Rengman, Hans Wilhelm, Herr, Uddevalla, Sweden

…and with a more strict structure:
Hans              Wilhelm          Rengman        Uddevalla       Sweden          Herr




Libraries seem to be far ahead both of archives – I think – and museums – I know – in looking at these challenges in a more systematic way.

The museum area has always discussed the classification issues with great emphasis but there seems to be an unknown something that leaves us a bit from the goal. It might be just a spell, or it might be worse – a professional blocking that makes us always to want the ultimate, all-covering and everlasting model to be created.

Libraries has a tradition going back to somewhere in the mid 19th century in subject classification and some archivists still tend to tell that there is no need for subject classification within the archives due to the special structure of the content – and some even say that there “is no need for the users as they always will have an archivist to ask…”

I still think that we have something to learn from libraries in getting the stuff out to the end users.  Museums seem to have a position somewhere between A and L in focusing so much on the issue – and yet we are not solving the problems.

While Libraries have their Dewey, UDC with international and academic acceptance and nationally adopted structures as the Swedish SAB which is common to almost all citizens and used in public and school libraries, museums tend to have a local system from one of two families.

On the one hand Cultural Heritage museums often stick to their traditional system, developed over years and trimmed to the physical shape of the local collection content.- A great tool as long as it don’t is any need for cooperation outside the museum.

In other cases, the museum can chose to adopt any of the intellectually developed systems or models that covers the entire world or a well defined aspect of it. AAT, ICONCLASS, ICOM COSTUME CLASS, OCM and a wide range of other systems like these.

The libraries also use the system of subject headings like the LCSH

Here we find other obstacles in collaboration. Either we can chose to use English language (British or American), or we can try to translate – finding that many of the elements are so bound to conceptual content with a cultural context and therefore not translatable. Some experiments in multilingual systems are known but to my knowledge no well-running ones. Leaves us with the idea of translate and modify for local need. One example is the OCM which were translated to Swedish in 1970ies and used for museum collections in general in Sweden, translated and adopted into other languages in Scandinavia and suddenly was found not comparable between Sweden, Norway and Iceland to take one of many examples.

I do not have a solution here! Still I want to raise this as something to be aware of in future planning.



SWEDEN – documentation history

In 1960-ies the first experimental projects on computerised museum documentation were launched at the Nordic museum and at The Swedish Museum of Natural History both in Stockholm.

During the 70ies the Skokloster Castle were restored and at that time a system was developed, that later on came to be a milestone in standardisation, and adopted by many regional museums. In the same period the OCM system were transformed to a Nordic, object based classification system.

In mid 1980-ies a working party proposed a new system-model called SAMOREG which took standardised thinking a step further.

1995-6 during the unemployment commitment SESAM we were a group of curators within the INSAM community at Nordic Museum, developing terminology and authority tools and a framework of rules like the Brittish Spectrum.

Over all, the development of computer systems based on this history has given us a bunch of home made systems in museums, and one of them has matured to a quite well-known and spread system, SOFIE. Within a group of museums the CARLOTTA system is developing, based on a bit unorthodox thinking – maybe the most interesting system in Sweden today. From the world outside we can notice that the Nordic Museum has taken PRIMUS, from Norsk folkemuseum in use and that the Modern Museum is using the TMS.

Today – when we are into another unemployment program – ACCESS – there is no requirement to use special software or to have the documentation based on a national agreement of basic information elements. Altogether – the above sketched history and outcome is, I believe -  due to a tradition with quite independent museums and a weak regulation on political level within the cultural heritage sector, we lack a bit of the national common standards thinking that we can see in Norway and Britain to mention a couple of interesting ALM examples.




National Museums of World Culture

Museum for Far Eastern Antiquities         
Mediterranean museum                           
Museum of Ethnography                         
Museum of World Culture


During the last two Years I have worked with a group of four experts from these four museums. The museums were quite recently merged into one organisation.

At the start we were investigating the needs and possible solutions for an authority-wide system for collection management. This was primarily based on the wish from the director and the steering group and we soon found that it was easier to say than to do. Collections numberings cover some 400.000 objects and the museum profile differs – two museums having ethnographical collections, while the other two can be categorised more as having a fine art profile, collecting items from the Far Eastern and from the Mediterranean areas.

The decision was taken, not to order a new system at that moment, but to spend a period of work within the database environment to find out whether the content was of the necessary quality or not and to find the points where standardisation was needed. There was also some other related questions to focus on.

The working group has during the period also met with other specialists in the museums and this has vitalised the dialogue between different groups of professionals.

Quite fast we found that there were three main fields where we could manage to get something done. By comparing and analyzing the content we found patterns which can be established as rules for further practical work within the museum. We can also wash out some terminology lists that can be adopted as organisation wide authorities – such as techniques and material. Based on these two types of results there is going to be better rules for writing and other manuals.

The discussion about the terminology lists is interesting as it raise the issue of end-user and target groups. The museums are by tradition focused on different user groups and also on persons with different language skills. One of the museums has almost all contacts in the international world of researchers and within a limited terminology context and thereby all documentation in English. One of them needs to handle Chinese signs and the level of knowledge is far above the normal levels of translation and transcription. The other museums have mainly Swedish documentation. The users are in the range from international experts to school children. Curators, Conservators etc also treat info in different ways. This is not a specific problem in itself – but connected to the other aspects of quality, it makes the total image quite complex when combining the four museums context fields.

We found out that for the element “material” we could go on three tracks. Try to make a list of terms that is close to everyday language and a tool for common end users. Collecting all actual terms in the museums and arrange them in a strict structure – a thesaurus. Leaving it all for single museums to decide.

Animaliskt material              
Ben, horn                            
Läder, skinn, päls                
Mänskliga kvarlevor            
Syntetiskt material/konstfiber                     

Ej bestämt                                               

Other animalic material
Human remains
Plant material
Syntetic material - other

Not classified


The result is that we agree on a very basic list of 20 words that gives the names for large or important groups that often are asked for. The list is not hierarchic and there can be an ambiguity in some cases. Different museums had different areas where they wanted. The museum with much objects of “mother of pearl” could not accept that this was just a part of “animalic” and so on.

Looking at the content in the dozen of databases within the NMWC we can see how both tradition, competence, awareness, and differing ways of meeting new technical demands, colours the way the content is described.







(gender form)                      
(gender form)                      
(“old” form – more valuable?)
(guest researcher)

(a house)                                                  
(a boat)

(a noun – the technique)

Some examples are needed to explain this. In using Swedish terminology hereI will just show the structure and some problems. I believe that it is obvious to you that this is not a way of saying that these museums have low quality. I have the clear view that one can go out and find similar examples in any group of selected heritage institutions.

The first two formats are equal and reflect the grammar gender in Swedish language. A common way to make retrieving of data easier is to have writing rules saying “use the t-form in all situations, to be consistent”.

The third is an older form, a synonym, that is today loaded with a value – this object is supposed to be of higher quality than a “målat” - or it might mean that there is flowers painted over the covering paint - but trying to stretch this in a discussion, we often ends up in laughter and red flushing faces as the arguments decomposes. It is overrepresented in museums documentation! It is obvious that this is not functioning when interacting with new Swedes or in the international community

“Redpainted” is a composed word that we do not want to have in a strict and standardised terminology. (At least not when discussing in this fundamentalistic perspective.) Still it’s there – used by many and understood by more. Some times even a part of even more complex terminology where “rödfärgad” is not just “painted in red” – it is painted with a special type of pigment and colour base used in Sweden – so “rödfärgad stuga” is a concept not equal to any red house…

“Färgad” seems to be another word – in many persons mind restricted to the meaning of dyed – the way you colour textiles by a dipping them into the pigment…

As seen from above the word “färgad” can have a wider meaning even on its own.

All these examples are subject to contextual prerequisites. Even if it, at the beginning, looks like there has been no or not sufficient proof reading, the second impression must be that

  • there is a lot more behind the single words than we can see at the first sight
  • personal bias makes a definition better or worse  - both writing and interpreting
  • research traditions differ and tend to invent language and terminology of their own
  • in house traditions strengthen different ways of writing
  • the context can be fuzzy and giving wrong signals

Two other fields can be used as examples.

Museum A uses a German based name convention in documentation – due to the research history in late 19th century while other museums use an English transliteration based on their traditions. This points us to two problems today.

  • The need for having a geographic master pointing to the “conceptual place” and a link for the German and English words, instead of just changing one of them and lose context and quality.
  • Whichever language selected, the quality of matching a name to Google Earth for example, is not at all good. GE uses a modern name convention while museums use the old one from context. GE uses an American English name form.

    (Is this a language issue or a political one?) Nevertheless  - today a ”non-answer” in Google is often equivalent to “non-existing”

The age dimension of quality is often forgotten. Many items have early documentation content which is outdated today – this is often the case in ethnographic or anthropologic collections. Many objects – considered strange by the early researcher are classified as “ceremonial” or “ritual” and coloured by our ethnocentric view. Even if not accepted today for humanistic reasons the terminology mirrors a viewpoint of its time and accepted as intellectual and research based knowledge of high degree.

(By definition this is valuable for its own sake in museum documentation. Maybe not useable, and a special effect of museum documentation compared to the rational information providing in image bureaus or research libraries, one can say).

  • We have a task here to both be correct and exact as of our knowledge today – meeting the needs from users of today – and at the same time show the history and development of research. Examples of this phenomenon can also be taken from geography names where country names changes with the political situation and just by using them you take a position in the war.

Conclusion from the two examples is that we should behave critical to the results of standardisation that is brought to us by systems that tend to set global standards – both large and small!

Relevant is also to consider the different end user groups, their skills and their bias.

This discussion has taken its start in museums documentation. I am convinced that it is transferable to Archives and Libraries world to some extent and for sure when comparing the three different heritage sources.

Looking at the National art museum in Riga, yesterday I was reminded of the international efforts on transliteration of names, actual both in museums and in libraries

Pierre Auguste Renoir

Alexander Rizzoni

1841 - 1919

1836 – 1902 

Pjers Ogists Renvärs

Aleksandrs Riconi



KMM is a joint project consortium with museums, universities and small companies. There is some core business dealing with infrastructure and common tools and then 14 projects targeted to specific purposes. KMM is also the Swedish node for the European Union Project MICHAEL+. KMM is a part of the ACCESS-project, and The Luleå University of Technology is the project owner from now on.

The KMM use the CIDOC CRM model as a ground for terminology and structure.

The task is to develop the services and open environments that collect, store and distribute the information that end users ask for.

    • Mjuk infrastruktur (xmpp/CIDOC CRM CORE, Interoperabilitet)
    • Games and Simulations in Education, Learning Environments
    • Regionala kulturarvsportaler (eTjänstemiljöer)
    • Kartsök
    • Digital Preservation (Centrum för långsiktigt digitalt bevarande, LDB)
    • Knowledge Management
    • Skolprojekt, pedagogik
    • Virkesstigen (Hemslöjd, pedagogik, Intangible Heritage)
    • Bonader
    • Bilddatabaser / genus
    • Museiarkiv (Uppordning)
    • Gustavianum (Digitalisering av samlingarna  och runstavar)

Infrastructure, standards, user support, connection to users experiences and more, are important.


    • KMM.ClassMaster (Klassifikationer, Terminologi “Subject Ontologies”)
    • KMM.NameMaster (Personer, Namn)
    • KMM.OrganisationMaster  (Organiationer)
    • KMM.GeoMaster (Platser, orter, regioner)
    • KMM.CollectionMaster (Michael plus)
    • KMM.TimeMaster (Mystisk än så länge)
    • KMM.Engine (Grundsystem)
    • KMM.Content (Datalagring)


In earlier museum projects, the classification schemes were looked upon with great interest. Many curators and information experts have tried to find out a way to compare classification systems and to have them merged. From my job with the Nordic outline I remember some of the museums trying to connect their specialised systems to the general one. It’s like just adding a cherry tree on the outermost branch of an apple tree – not fruitful at all. Instead it’s better to see that there is need for both of them. One general – that might just classify the entire collection under the same code (tram traffic, archaeology) and then use the branch specific system – developed by branch specialists – to divide the content within the systems.

Another experience is that when comparing one system for clothes – starting with men/ woman – over/under the waist – with another – beginning in holidays/everyday use - outer layer/inner layer you can of course find points where a dress is placed in these systems. But – when moving around (deeper, narrower and sideways) you will find by nature, that the context is diverging again. In other cases the two systems should be the same system.

One has to be aware of the prerequisites and the context for each of the systems to have them making sense.

In KMM there is a challenge to test these issues in a more offensive way. Two main reasons are discussed besides the practical goods of having the systems running in an environment where many museums help to develop the content quality.

  • To have a general technical system developed – suitable to handle as many authority systems as you wish and to provide them to the users.
    • A general resource for categorisation and documentation with authority lists, thesaurus etc.
    • A general tool for building such authorities
    • A way to force – or help - the user to the high level of standardisation in input phase, and to lead the user around any free format traps that can lead to later difficulties.
  • To have a test bed for looking at the possibilities of decision support and expert systems within the field of museum classification- Is it possible to make automatic classifications – and to what extent?
    • A study object for the field of automated services and expert systems use in museums material and workflow.

And – as a result – an answer if it is possible to build such decisions and support into a workflow system within the heritage sector?


As well as this research and experiments will show a standardised way of handling more and more of the actual data and lead curators through the process towards a higher degree of quality in the sources, we may also be aware that this can have the opposite effect.

Case 1
When we use a standard terminology we provide the curator with a set of words to choose from. If there aren’t any suitable alternatives, you are stuck in the same situation as the child trying to put small pieces of different shape into a box with holes - round, square, triangular - and it comes a star. You end up in fetching a saw to trim the star or trim the box – Guess what’s the best?
(If you get just one star, it’s easier to trim the star to a round – if you will get many stars, it will be needed to trim the box to swallow all of them.)

You risk to be forced to define an object as belonging to another category – to be something else than it is.

Case 2
In a case where You have quite good information and detailed knowledge and the list of terms to use is just built on words or concepts at a basic level, you will risk to be forced to declare an object belonging to a more general group than your expertise can find out. If there is a “Windsor chair” and You do not have it in your list, You look at the level above if there are “chairs”, If not You climb another step and decide that the object is a “furniture” – which is indeed all right, but not precise enough.

You have lost a lot of knowledge!

Case 3
From the other side of view, the end users, we can see a similar risk. The question put to the database might be so simplified or out of focus that the answer is either a mix of relevant and irrelevant objects, or a far too big, and wide, group of objects to be interesting. We can also see the general risks over time of having forwarded a heritage nomenclature or a language domain , simplified to that extent that is not usable for more complex purposes.

…and this goes for both the end user and the professional staff that is leaning too much towards the tools of standardisation.

By pointing at all these risks I will take the standpoint that this experiments are very interesting to carry out, but that we might have as much to gain by study the effects of these tools and routines, as we win in practical support in the first stage.




So – how can we fit these two projects together and how can we make benefit from them to the community.

The terminology comparison and the analysis of the content in a database or in a group of databases can give us direct knowledge to syntactic and structural handling of the language. It can also give us a higher degree of awareness of methodological approaches to the used language and help us making better rules for future work. There can be openings for looking at museum documentation terminology in a research perspective at any of our museology universities.

Working in a project like KMM gives the opportunities to use a large amount of real data from different environments for testing. The group of museums in the project provides both a wide range of objects (=terminology) and a varying set of user history, in the sense of different habits and writing rules. The merging of these data, to clean it all up and to make it searchable in a common tool is of course an expected outcome. Furthermore we will be able to see even wider structural patterns in the material and thereby to be able to develop terminology rules in a more general community perspective.

Considering standardisation as a tool of importance, and cleaning and maintaining of databases as necessary work we should not forget the human aspect of the content in our documentation.

Besides being labels and handles for getting in touch with the content in a structured way we also have to face the fact that the information is a message about the context in itself, the language has a cultural level that has to be documented in itself and it all shows for the future a snapshot of the time of origin – both ancient cataloguing and today’s electronic documentation.

There is need for two levels – or two groups of terminology. We must keep the difference between the structured and formalised terminology that is needed for comparing and exchange and the free forms of language that bears the time, the person and the flavour of history in…

In a large scale the amount of objects in a group or class will be unmanageable. It might point to a revision of the classes and subclasses. If so done – there is a risk that we construct patterns and syntax not yet seen. If not – we will have to stick to very large groups of items and data base posts that carry a similarity that hide the originality of the object.

By studies and research, by discussing methods and comparing results we can share experiences an rise the level of  meta-knowledge in this field. Here I see many interesting fields of comparison not only within museums, but also in the entire heritage sector where we all can gain from earlier results and experiences



What’s so good about standardisation? Well – As I can see it, Standardisation will be the tool for us to make contacts outside the institutional box possible. To be able to connect to other museums, to meet between archives, libraries and museums and to support the end users in having easy access to the data. On the other hand, a believe in the magic of standardisation makes us prisoners in the system, peeling away uncomfortable odd data and drives us to use more square boxes than necessary.

The NMWC studies have put a finger on the concrete examples and their background. Systems for catching multi-institutional data must be constructed in a way that do not cut away one aspect on behalf of the others.

The KMM project gives us a possible tool to large scale manipulation of information and a platform to discover ways of rational data handling. The risks in driving the standardising too far will be noticed.

The goal should be to keep the natural language as untouched as possible, to adapt standard terminology as tools and keys “outside” the documentation in itself and to try to hang on as many tags or information viewpoints as possible – a challenge both to the processes and the professional role.

We need discussion and we need research within these fields and we can benefit from looking at each others systems with new eyes within the entire ALM and heritage sector.

+46 (0)70 718 23 25

Skype Me!