Conclusion: Patterns of Non-Territorial Autonomy

Abstract

This article seeks to generalise about the significance of non-territorial autonomy as a mechanism for the management of ethnic conflict on the basis of a set of case studies covering the Ottoman empire and its successor states, the Habsburg monarchy, the Jewish minorities of Europe, interwar Estonia, contemporary Belgium, and two indigenous peoples, the Sámi in Norway and Maori in New Zealand. It begins by assessing the extent to which the spatial distribution of ethnonational communities determined the range of autonomy options available—whether these might be territorial or whether only non-territorial autonomy would be realistic. The article continues with an assessment of the significance of ‘autonomy’ in circumstances where the institutions with which it is associated enjoy a non-territorial rather than a territorial writ. It concludes by suggesting that in almost all cases where autonomy is extended to a minority within a state this is exercised on a territorial basis, and that in many cases of non-territorial autonomy, or national–cultural autonomy, the powers assumed by the ‘autonomous’ institutions are substantially symbolic. It argues that notwithstanding the limited empirical evidence for the existence of non-territorial autonomy, this device should not be written off at a normative level.

Acknowledgements

The author is grateful to Karen Barkey, Rainer Bauböck, Karl Cordell, Richard Hill and Per Selle for comments on an earlier draft.

Notes

1. The Ottoman data use 33 primary units: 26 formally designated as vilayets, five special districts (one formally designated as a sanjak), and the city and suburbs of Istanbul as two separate units. The Russian data include 71 guberniĭ and 18 territories, mainly in Asia, formally designated oblast (region). The 17 crownlands of the ‘Cisleithanian’ portion of the Habsburg monarchy, which functioned in the manner of the component units of a federation, had their origins in complex historical developments, varied greatly in size, and had formally different designations (such as archduchy or duchy, though three had more elevated titles: the Kingdom of Bohemia, the Kingdom of Galicia and Ludomeria, as Austrian Poland was named, and the Kingdom of Dalmatia).

2. This ignores the significant Slovene population lying outside the borders of Yugoslavia (or the Kingdom of Serbs, Croats, and Slovenes, to give it its official name), notably in the Austrian province of Carinthia.

3. Expressing the Sámi parliament electorate as a proportion of the general electorate for Storting elections in 2013 would give a lower estimate, 0.4%; by this measure, Sámi account for about 4% of their traditional territories in northern Norway; derived from Statistics Norway (2015). The Sámi area is defined as the two most northerly Norwegian counties plus the northern part of a third county. It forms part of a broader area, Sápmi (the country of the Sámi, formerly known as Lapland), which extends over all of northern Norway, much of northern Sweden and Finland, and Russia's Kola peninsula (see map in Slaastad, 2014, p. 7).

4. The questions were: ‘which ethnic group do you belong to?’, with such options as New Zealand European, Maori, and so on, and ‘are you descended from a Maori (that is, did you have a Maori birth parent, grandparent or great-grandparent, etc)?’. Of those including ‘Maori’ in response to the ethnic affiliation question, 41% opted for Maori only, 49% for Maori and European and the remainder for other combinations; derived from Statistics New Zealand (2015).

5. Computed from Statistics Norway (2015), and from data on the Sámi electorate supplied by Statistics Norway.

6. Stalin (1948 [1913], pp. 26–35) spent almost as much time attacking Bauer as attacking Renner, even though the Bauer text that was his target simply offered a detailed description and endorsement of the Renner scheme (Bauer, 2000 [1924], pp. 281–292). This may have arisen because of the intellectual weight attached to Bauer as a theorist of nationalism, and seems to have given rise to the later misleading description of the Renner model as the Renner-Bauer one.