The initial draft of this document was created in April 2016 by a group tasked with devising the policy on open access and open data for the Árni Magnússon Institute for Icelandic Studies. The group comprised Haukur Þorgeirsson, Ólöf Benediktsdóttir and Steinþór Steingrímsson.
The document established objectives with regard to open access and open licensing for data generated by the institute. Likewise, the document attempts to define the concepts of open access and open data and how to differentiate between the two for the purposes of the institute.
The document was amended in 2019 and 2020 by Steinþór Steingrímsson and Haukur Þorgeirsson. The policy was subsequently approved at the institute's staff conference on June 9, 2020.
Note on Copyright & Terms of Usage
All decisions with respect to licensing must be made in consultation with the relevant rights holders in cases of copyright, for both institute employees and parties outside the institute.
In certain applications, copyright attribution may be problematic for databases and institute manuscripts. In some cases copyright applies to a portion of a manuscript or database but not to other parts. For example, in a database copyright may apply to a single entry or to how the data is collated. And under some circumstances it may be necessary to obtain a legal opinion to determine whether copyright applies to certain data.
Efforts shall be made to use Creative Commons (CC) licenses whenever possible. The advantage of such licenses is that they are widely used and presented in plain language so that users are not required to interpret complicated rules. There are several types of CC licenses. The most accommodating license allows all usage as long as the source is credited. This means that anyone could access the data, copy them and sell them if they so choose. They could also change the data and distribute an altered version, use it for research purposes or for any other purposes they see fit. Other CC licenses limit the scope of usage in a number of ways. For example, some prohibit all usage for commercial purposes. Some prohibit any changes to or omissions of the data. Some require that all derivatives carry the same license attached to the original data.
CC licenses are used for data and data sets, but also for images, multimedia materials and journal articles. Stjórnmál og stjórnsýsla (the Icelandic Review of Politics & Administration), the journal of the Institute of Public Administration & Politics at the University of Iceland, publishes all of its articles with a CC-BY 4.0 license.
The following licenses are applicable:
The user must credit the original source of the material but is otherwise free to use the material however they see fit.
The user must credit the original source of the material and may use the material as they see fit, but all derivative work must be licensed under the same terms as the original material.
The user must credit the original source of the material. They may redistribute the material for commercial and noncommercial purposes, but only if the material remains unchanged and in whole.
The user must credit the original source of the material. They may use the material however they see fit but not for commercial purposes.
The user must credit the original source of the material. They may use the material as they see fit, but all derivative work must be licensed under the same terms as the original material. The material may not be used for commercial purposes.
The most restrictive license. The user must credit the original source of the material. They may not use the material for commercial purposes and may only distribute the material unchanged and in whole.
More information here: https://creativecommons.org/licenses/
A number of museums, galleries, libraries and archives have released and permitted usage of some of their data with open licenses. The following is a list of institutions that use CC licenses to some degree. They include the National Gallery of Denmark, the National Library of Sweden, several Finnish libraries, the British Library, Rijksmuseum in Amsterdam, the Museum of Modern Art in New York and others: https://wiki.creativecommons.org/wiki/GLAM
It is also becoming increasingly popular for academic journals and conference proceedings to make articles available with open licenses. In addition to Stjórnmál og stjórnsýsla (the Icelandic Review of Politics and Administration) mentioned above, two other Icelandic academic journals are also published with CC licenses: Tímarit um uppeldi og menntun (the Icelandic Journal of Education) and Tímarit um viðskipti og efnahagsmál (Research in Applied Business & Economics). Foreign journals include Oxford Open, a publication put out by Oxford University Press with open licenses.
Open access means that anyone can interface with data or read articles through open web access.
The concept of open access entails access without limitations. In some settings it has become popular to delay open access, i.e., to provide open access only after a certain amount of time has passed after publication. This practice, however, may reduce the material's exposure and, in turn, its impact and how often it is cited.
The Árni Magnússon Institute publishes two journals: Orð og tunga and Gripla. To facilitate the inclusion of the publications in international databases, thereby increasing visibility, it is essential that the articles are as accessible as possible, that they include all necessary metadata and that they are assigned a DOI number. To this end, the material must be provided with open accessibility to satisfy the established criteria. We propose that the journals be published online using the Open Journal System (OJS), which is already standard practice at several other institutions associated with the University of Iceland, and which may serve as a model for our own system.
The institute's journals shall endeavor to publish articles with CC licenses. The editorial staff of the journals shall see to the precise implementation, but information for authors shall be available on the journal websites.
When possible, older material shall be made available online and assigned a DOI number. In these cases it is not necessary to apply CC license to the articles, but the article authors may need to give their consent for the materials to be released online.
General guidelines shall be put in place with regard to the online publication of the institute's books. Books published by the institute may, in some cases, be released online two to three years after the print editions are published. The details and practice of these online publications must be reviewed by the institute's publication committee. In consultation with authors/publishers, the publication committee may also undertake the release of older books with open access.
The institute recommends that institute staff ensure that they may release their peer-reviewed articles with open access after initial publication. This applies to journals other than the institute's journals. In some cases this may prove unfeasible, and it is understandable that authors may need to accept other terms, especially in cases of publication in widely read journals.
The institute's website includes a list of staff publications with details on all peer-reviewed articles. The webmaster sees to it that the staff regularly updates their publications.
Staff may also opt to list references to work that has not been peer-reviewed as well as lectures. This list should be maintained separately but accessible from the list of peer-reviewed work.
Staff shall be able to add their articles through a dedicated system. The University of Iceland's "Opin vísindi" CRIS may be a suitable system for these purposes. Otherwise the web committee shall identify other resources to address the need.
In some academic disciplines it is becoming more popular to provide open access to research data, e.g., datasets used in published research. For example, the suggestion has been made at recent LREC events and other conferences on language technology that data used in research presented at the conference be made available online with open access. That said, it is widely understood that open access is not feasible in all cases. Nonetheless, the institute must make provisions for this kind of access and provide the support necessary to carry it out.
Access is already open for most of the institute's online databases. They are readily searchable for public use. There are, however, exceptions, the reasons for which vary from database to database.
The institute endeavors to make all of its databases searchable online, unless special circumstances prevent this arrangement. The institute's website shall still present the existence of those databases not searchable online so that outside users may at least request information when necessary.
Objectives for Open Access
- Usage licenses and terms shall be established for all databases and articles
- Articles from the institute's journals shall be published with open access
- It shall recommended that staff also make their other own peer-reviewed articles available online with open access
- A database shall be maintained on the institute's website detailing peer-reviewed articles authored by staff
- All databases shall be made readily searchable online unless prevented by special circumstances
- The institute's website shall include a listing of all databases, their level of accessibility, if they have open access, and rationale provided for any limitations to access
- Accessibility for all databases shall be reviewed
Open data entails not just that the public is provided with access to data, but also that anyone may use, transform and redistribute the data for any purpose. The objective is to maximize access and usage of the data. That said, the concept of open data may still apply even when limitations have been imposed through the usage license.
When data are readily accessible online without any usage terms or licenses, it is generally understood that the data remain the property of the party that created them and that all rights are reserved. These days, however, data are increasingly made available with a relevant usage license. This is becoming standard practice for both public institutions and private companies around the world, especially when they engage in research. Generally, institutions and companies apply an existing license. For example, one of the Creative Common (CC) licenses.
Below is the data publication pyramid from the Leru Roadmap for Research Data. It presents the various layers of data in a way that is presumably a decent representation of our own institute's data.
What Kind of Data?
All data that the institute makes accessible online will need to be assigned a clear usage license. It is preferable to use internationally recognized licenses like those from CC. The terms of licenses should also be easily comprehensible for users if they are not written expressly for a specific database belonging to the institute. The terms of licenses should require neither specialist legal knowledge to interpret nor a certified translator to translate.
Different data may require different licenses. In some cases it will not be necessary to limit usage in any way since the objective is to maximize data usage.
For example, in recent year various language resource data sets have been made available in this fashion, including the Pronunciation Dictionary for Icelandic as well as text corpora like the Saga Corpus, the Hjal Corpus, the Málrómur Corpus and a corpus with integrated sound and text files of parliamentary discussion in Iceland. While all of the above data sets are available with completely open licenses, most other Icelandic language resources have required some kind of usage limitation.
Our purposes primarily pertain to existing databases and existing data. In addition, there are research data associated with published studies. There may also be data associated with preliminary research, which may be of use and would require little effort to make accessible.
Each department shall review the data it has already made available and determine what type of usage license ought to apply. If there is no usage license in place, then an appropriate license shall be selected for the data in question. Each department shall also determine what other material may be made available and how best to handle various types of data.
What Kind of License Should Apply?
According to the institute's policy, a completely open license, e.g., CC BY 4.0, shall apply to all the institute's data that are already accessible from outside the institute, unless there is a compelling reason to limit accessibility. Any limitations shall be accounted for and the rationale shall be presented wherever the data are distributed.
Objectives for Open Data
- Usage licenses shall be determined and/or reviewed for all institute data available online. (A review may be necessary for data that already carry a specific usage license, if only to provide a rationale for any limitations.)
- If the usage license in question imposes any limitations then the rationale for those limitations shall be presented on a case-by-case basis.
- The institute's departments shall review the usage terms for data they make available for public use.
- The institute’s departments shall review whether it is possible to make other data/databases available for public use.