- Disputes over territory are a major contributing factor to the disruption of international relations. We believe that a cumulative, integrated, and continuously updated resource providing information about such disputes in an easily accessible form would be of benefit to intelligence analysts, military strategists, political scientists, and also to historians and others concerned with international disputes. We propose an ontologybased strategy for creating such a resource. The resource will contain information about territorial disputes, arguments for and against claims pertaining to sovereignty, proffered evidence for such claims, political and military motives (overt or hidden), and associated conflicts. Our approach is designed to address several issues surrounding the representation of geopolitical conflict, including the tracking and individuation of disputes and the validation of disseminated information.
In what follows, our focus is on the ontological representation of territorial disputes and the different sorts of entities associated therewith. The ontology is intended to be used for the consistent annotation of data and information about territorial claims, arguments for and against such claims, political and military motives (overt or hidden), and the different types of conflicts associated with territorial disputes, from occasional skirmishes and limited engagements to terrorist campaigns and outright war.
We begin with a survey of some of the problems faced in representing territorial disputes in data and information systems. Second, we sketch how Basic Formal Ontology (BFO) 2.0 can be used as a framework for a realist understanding of different ways in which individual and group agents participate in territorial and other conflicts (http://www.ifomis.uni-saarland.de/bfo/). Third, we provide a framework for the handling of data about territorial claims; such claims can cause problems for a realist ontology, since they are often marked by the use of empty or disputed reference, for example when opposing parties in a territorial conflict produce maps of putative political entities in a given territory which cannot simultaneously be veridical. Finally, we propose a way to capture the relationships between motives and arguments underlying territorial claims, and we conclude with a case study of a territorial dispute between Japan and
II.
representation of territorial disputes include:
ontological
(
militias, liberation movements); leaders and representatives of the disputing parties (politicians, terrorist leaders, tribal leaders, religious leaders…).
as the UN, NGOs, aid agencies; mercenaries, religious bodies, press and broadcasting agencies, salient minorities (for example, Russians in East Ukraine).
such as diplomatic notes, treaties, claims, maps.
The first step toward the representation of a territorial dispute is the ability to identify the dispute itself. This can be problematic, since how we should individuate any given dispute may be unclear. For example, is the on-going dispute over oil resources in the Arctic a single dispute or an interconnected web of disputes, involving multiple different countries? (See Supplementary Material Figure 4)
need to take account of conflicting descriptions of both the dispute and the surrounding matters of (actual and purported) fact. For instance, the People’s Republic of China claims the island of Taiwan as one of its provinces; the inhabitants of
China, which was (they allege) wrongly forced off the mainland in 1949. Such incompatibilities cause problems for theories – like the theory defended by John Searle – which see political entities such as nations as social objects maintained in existence by the beliefs of the parties involved. An approach like Searle’s, it seems, can provide an account of the ontology of geopolitical entities that is satisfied only in a world in which territorial disputes do not exist [1].
Because territorial disputes do indeed exist, as also do the associated conflicting claims, an ontology of territorial disputes must find some way to do justice to the fact that given claims may have no referent in the real world. To capture, for example, what is involved when disputants talk about “the Chinese territory of Taiwan” or “the Taiwanese territory of China,” or when a neutral observer talks about the disputed territory of Taiwan combined with (the rest of) China, we need to find a way to link claims to the corresponding geographical regions without also prejudicing our representation in favor of one or other party.
Something similar holds when claims issued in the course of a dispute come in the form of accounts of a nation’s historical presence in a territory which are offered as evidence to legitimize a claim of present sovereignty over that region (for example, accounts of the Jewish presence in the territory of present-day Israel since the time of the First Temple). Again, our ontology would need to be able to represent the content of such reports without necessarily endorsing their claim to truth, and the same would hold of geographic or oceanographic reports documenting measurements of the boundaries of given territories in ways that may affect claimsof sovereignty. For example, Russia has claimed approximately half of the Arctic Ocean (1.2 million square kilometers) on the basis of measurements of its continental shelf (the natural prolongation of its landmass, as defined by the United Nations Convention of the Law of the Sea (UNCLOS) – see [2]). The international community has greeted these measurements with considerable skepticism. Nonetheless, it is essential that analysts be able to search for data about the region Russia does claim. How, then, do we make sense of the content of our words without thereby imparting to our words a referent in the world? In what follows, we propose a solution to this problem, whose goal is to render an ontology capable of handling both true and false claims made within a territorial dispute.
IV. ENTITIES IN A TERRITORIAL DISPUTE
One of the questions an ontology is designed to answer is: What kind of entity is X ? Consider for example the role of being an arbitrator of a dispute. In BFO terms, this is a specifically dependent entity – it could not exist apart from the person who bears the role. Or consider the information about some arbitrator contained in some document; to what kind of entity are we referring when we refer to this information? From the BFO point of view, an information artifact of this sort is a generically dependent entity, which means that whereas it requires some bearer – for instance, some hard drive – in order to exist, it does not require any specific bearer, because it can be copied from one hard drive to another; forwarded over email, or printed out in the form of a paper document (http://code.google.com/p/information-artifact-ontology/).
In addition, an information artifact is a kind of entity that is capable of being about other entities. Territorial claims, maps, treaties—all exist as information artifacts in this sense. They are, like persons and roles, territories and territorial boundaries, what BFO calls continuant entities, which means that they continue to exist through time, even while undergoing changes of various sorts, for example in the form of amendments, codicils, and ratifications.
Territorial disputes, in contrast, are occurrent entities, which means that they are entities that occur in time and unfold themselves in successive temporal parts. Territorial disputes will differ along a number of dimensions, including their duration, the parties involved, and the degree to which they involve different levels of violent conflict.
They will also differ according to the territories to which they relate, the boundaries of these territories, the populations of human beings occupying these territories, and so on. An ontology to support reasoning with territorial dispute data will thus require a resource such as the I2WD Geospatial Ontology (http://milportal.org), incorporating also references to the different kinds of fiat geopolitical entities described in [3].
In the case of claims that nations stake upon disputed territory, the information artifacts involved may be descriptive or directive. For instance, during the Six Day War in 1967 Israeli forces seized East Jerusalem and asserted that this region is (and always had been) part of the geopolitical region of Israel. Thus, Israel’s claim on East Jerusalem is formulated as a piece of descriptive information. However, this claim was associated with directive information specifying how persons should conduct themselves with respect to that territory, declaring inhabitants of East Jerusalem subject to Israeli law and restricting access to the region by non-Israelis.
Territorial claims are often bolstered by arguments aimed at establishing their truth. For instance, to justify the seizure of Palestinian territories, Geula Cohen of the Israeli Parliament argued in 1999, “The Jews did not come back to Israel to be safe but to build a nation on the lands given to us by the Bible.” [4] In this statement, Cohen gives an argument in support of Israel’s territorial claim on the basis of divine right. Here, we can distinguish Israel’s territorial claim, Cohen’s argument in favor of this claim, and the religious beliefs underpinning this argument, all of which are salient to representing the territorial dispute as a whole. Or consider also the 1994 argument of Stephen N. Schwebel, in favor of the legality of the Israeli settlements on the basis of the principle of a sovereignty vacuum [5] (See Supplementary Material Figure 5). Another important feature of territorial disputes are the motives of the disputing parties. Salient motives include: total autonomy, independence, or secession (for example of Quebec from Canada, of Scotland from the United Kingdom, of Catalonia from Spain); local autonomy (of Kashmir, South Tyrol, Sicily) within one or another existing sovereign nation; economic advantage via the exploitation of natural resources (for example petroleum and natural gas in the areas surrounding Hans Island and Paracel Islands, and in the Aegean Sea); restoration of territory held to have been lost (of Gibraltar to Spain, of Ceuta to Morocco, of Belize to Guatemala); expansion of territory (of Canada, Russia, the United States, Denmark, and Norway into the Arctic Circle); strategic gain (of Tuzla Island and the Strait of Kerch for Russia).
This collection of data relating to motives is complicated by the fact that overt motives may differ significantly from hidden ones. We contend in what follows that an ontology of territorial disputes should have the resources to represent motives of both types, though we recognize that obvious problems arise in regard to the latter since it is possible for the motive that is driving a territorial claim to be perfectly concealed. This however is not different in principle from what is involved when gaps in scientific knowledge are revealed by new discoveries.
Persons, organizations, and governments engage in a wide array of disputes over plans, goals, predictions, decisions, policies, laws, beliefs, property, and territory. Additionally, they engage in a similarly wide array of conflicts. But what kind of things are disputes and conflicts, and how do they relate to entities of other sorts?
As a provisional characterization, both disputes and conflicts – whether or not they are territorial in nature – are relational processes in BFO terms; thus they are processes dependent upon and involving as participants at least two agents. In the case of a dispute, the participants have views that clash, and the dispute unfolds in a series of sub-processes in which each disputant offers claims and arguments in the hope of convincing others of the correctness of their views. A conflict, on the other hand, is a relational process that is made up of inherently hostile (violent) interactions among participants. Not every dispute engenders conflict: a disagreement over the precise demarcation of a border, for example, may be resolved through negotiation. And not every conflict involves a dispute, as when one person strikes another in some random attack.
When two agents – which may be either single persons or more or less formally organized groups of persons – are involved in a dispute, then there exists also a relational quality, which in BFO terms is a specifically dependent continuant that inheres in multiple bearers and which connects them together. We can refer to this relational quality as a state of dispute. A state of dispute comes into existence at a certain time (for example as the result of an act by one of the parties of staking or contesting a territorial claim); but it exists thereafter in its own right until, perhaps through some further act, it goes out of existence. A state of dispute is essentially relational; thus it not reducible to non-relational qualities inhering separately in the involved parties (compare in this respect relational qualities such as claims and obligations).
Once the state of dispute comes into existence and it
preserves its identity for as long as it exists even while
undergoing a variety of different sorts of changes (for example
in intensity of associated conflict). Even while we do not
believe that it is possible to articulate an exact account of the
identity conditions for disputes over time, we nonetheless
believe that it is in many cases unproblematic to identify a
given state of dispute as one and the same from one time to the
next. Just as an obligation comes into being upon the making
of a promise – for example as documented in a written contract
– and only ceases to exist upon either the fulfillment of the
promise or the waiving of the obligation by the one to whom
the promise was made, so a dispute comes into being upon the
act of instigation and only ceases to exist upon either (
Consider, now, the special case of territorial disputes, a typical case of which involves two or more governments in dispute over sovereignty in regard to some specific territory. In our view, the state of dispute is an entity existing through time as an entity in its own right inhering in the participant agents. The state of dispute begins to exist because of actions on the part of one of the parties involved (for instance, Nation A moves armed forces into a territory claimed by Nation B; Nation A releases a map that depicts a region that is claimed by B as falling within the geopolitical boundaries of A). Dispute processes may then ensue, for example on the diplomatic level, but the state of dispute exists even during periods of time when no such processes are occurring.
In some cases, such dispute processes lead to outright conflict, and there then arises a new relational quality called a state of conflict, with subtypes including a state of war, between them. Both states of dispute and states of conflict are relational qualities that inhere in two or more agents.
VI.
A. Kinds of Data
We turn now to the handling of data about territorial
claims. Most of these data are readily treated with the resources
of the I2WD ontology framework, including the Information
Artifact Ontology, Geospatial Ontology, Time Ontology, Event
Ontology, and Agent Ontology, with BFO as foundation. This
suite of ontologies can be used as a tool for tagging different
types of data salient to territorial disputes, including:
(
(
(
(
B. False and Disputed Information
Any territorial dispute will involve conflicting information about the status of some territory, the location of the relevant borders, the soundness of the arguments supporting territorial claims, the (descriptive or rhetorical) character of given press releases, the validity of existing treaties and rights of other nations or groups, the interpretation of salient judgments of international law, the history of the disputed region, and so forth. Sometimes the salient information can be classified either as true or false simpliciter – for example, in the case of geographical coordinates of given landmarks. In most cases, however, we shall need to refer to claims as true or false in the eyes of one or other of the disputed parties, or as being such that their truth or falsehood is uncertain (with various modalities). We now suggest a way of tagging information along these lines, distinguishing three categories of information that fall short of being true simpliciter: a. Information that has a truth-value that is to a degree uncertain. b. Information that is not false, but has some related defect, for instance, in being misleading. c. Information that falsely asserts that a relation or a particular exists when it does not.
Categories a. and b., which cover many territorial claims, can easily be handled within our framework, For instance, it is (currently) uncertain whether Hans Island belongs (or should belong) to Canada or to Denmark, whether Bethlehem belongs (or should belong) to Israel or to Palestine, and whether the Paracel Islands belong (or should belong) to China, to Taiwan, or to Vietnam. Confidence in such claims begins as a cognitive process of assessment that has as input, the claim, and, as output some degree of confidence (uncertain, very uncertain, and so on) that will be used to tag the information in our knowledgebase. Information in category b. that is not false, but in the vicinity thereof – because it is metaphorical, bullshit, rhetorically embroidered, evokes codes only understood by its intended audience, and so on – can be handled by tagging the claim as output of one or other kind of performative act (of misleading, provoking, inciting, and so on). In this way, we can draw attention to the fact that the information is being communicated with a special purpose or in a special context that modifies the literal meaning of the words being used.
However, category c. cannot be dealt with so easily. In many cases, analysts do know with a high degree of certainty that a claim is false. For example, in III we noted that Russia has claimed 1.2 million square kilometers of the Arctic Ocean on the basis of a false report of oceanographic measurements of its continental shelf. Our ontology must be able to represent what that report is about, in this case, the 1.2 million square miles that (Russia claims) comprise their continental shelf. The problem is that, on the supposition that the Russian claim is false, there is no such entity as the Russian continental shelf of 1.2 million square kilometers.
A common strategy for representing false statements involves employing reified RDF triples, where an individual RDF statement may be annotated with the quality “false.” This strategy allows knowledge about an RDF triple to be expressed in two steps: the first consists in representing the triple by an instance of a statement that has subject, predicate, and object indicated separately in three different triples. The second step involves creating assertions about that instance as if it is a statement – in our case, an assertion of falsehood. This strategy allows for making statements about statements, but it has largely been found to be inefficient by many users, who find that it dramatically increases the run-time of queries – often making them impossible (though this may change in the future with the introduction of new strategies [6].
We are exploring an alternative two-step approach that begins by appealing to the family of lacks relations introduced in [7] in the context of a treatment of negative assertions concerning medical documents. For example, the proposed relation lacks_part would hold between a particular p and a universal U whenever p has no instance of U as part (such assertions will be made where there is an assumption that p should have or is expected to have a part of this sort, as for example in: John is missing his left arm. Since the particular (John) and the universal (left arm) both exist, the assertion of a lacks_part relation between them is perfectly in order from a realist point of view. When applied to territorial disputes, this strategy would allow us to posit the fact that corresponds to a false claim: e.g. that the Russian continental shelf lacks an extension covering 1.2 million square miles. We can then tag the false claims as being both false and also about the corresponding lacks relation. This allows us to interpret the class of false information content entities as bearing an is about relation to an existing portion of reality. This allows us to deal with false believes held for example by specific governmental organizations while remaining in conformity with the principles of ontological realism.
VII.
ARGUMENTS, MOTIVES, OBJECTIVES
Another significant component of territorial claims are the arguments made by disputing parties in defense of their claims. Any adequate representation of territorial disputes needs to capture the arguments for or against the truth of given claims, together with the other information content entities delineated above. To this end, we likewise treat arguments as information content entities borne by documents of a range of different sorts. Like claims, arguments are generically dependent continuants that can exist in many bearers, for example in multiple hard drives, in printed form in newspapers, in fliers posted on walls, and so forth.
We should think of arguments as wholes that are comprised of informational parts. For example, the argument supporting the thesis that Russia exercises sovereignty over half the Arctic Ocean involves both the conclusion that Russia exercises sovereignty over a particular region and other claims (the premises of the argument) offered to support that conclusion, here: that international law stipulates the criteria for a nation’s continental shelf, that the measurements of Russia’s continental shelf include half the Arctic Ocean, and that whatever falls within a nation’s continental shelf belongs to that nation. The argument in favor of Russia’s sovereignty over half the Arctic is the logically ordered collection of these information artifacts. Dividing an argument into its parts allows our ontology to keep track of the ways arguments are amended over time. In the first place, we can tag the claim that Russia exercises sovereignty over the Arctic as the conclusion of an argument. We can then situate that claim within an aggregate of other relevant claims. If different reasons are given at a different time, then we can treat those supporting reasons as comprising a separate argument on behalf of the same conclusion. Finally, if some argument is bolstered or diminished by new evidence (e.g., if an independent party issues the results of new and more precise measurements of Russia’s continental shelf) then we can represent that emendation.
Next, we consider the various kinds of arguments that could be offered in support of different kinds of claims. Kinds of argument are differentiated by what kinds of considerations they appeal to, for example evidence from geography, from geology, from history, and so on. Brian Sumner [8] identifies nine such kinds of considerations nations might appeal to in defense of a given territorial claim:
(
(
(
(
(
(
(
(
This classification becomes especially useful when we analyze arguments employed in territorial disputes from the perspective of the known or suspected motives of the governments involved. For instance, there is reason to believe that the Arctic region is the site of untapped natural resources. Russia’s stated arguments in favor of its sovereignty over half this region may make no mention of these resources, and yet an analyst can reasonably suppose that the intention to exploit those resources is one of Russia’s motives for claiming sovereignty. Thus, we first propose distinguishing between stated arguments and known or suspected motives, and then viewing arguments in light of these motives. Motives comprise the objectives the government has in winning the territorial dispute, whereas arguments are devices to facilitate progress toward gaining these objectives. That is to say, arguments are only one part of a government’s plan to realize its objective, which is authority or sovereignty over a given disputed territory.
This objective will in every case be embedded within a nation’s efforts to realize broader economic, political, ideological, and military goals. A country might have a plan whose objective is to grow its economy, and this plan might include subplans for some sort of political or military action to achieve sovereignty over some region and exploit its resources. ([9] provides a detailed discussion of some Norwegian, Russian, and Finnish strategies pertaining to territorial claims in the Arctic along these lines.) The components of these subplans will in turn involve, at still lower levels, plans concerning how to achieve this task, whether by vigorously defending some claim at the United Nations or intimidating the military craft of other nations in a given area. VIII. APPLICATIONS: THE CASE OF THE KURIL DISPUTE We will now apply our ontological approach to data about a specific dispute, namely that between Russia and Japan over the Kuril Islands, which has been on-going for over a century (Bobic, 2012, see Figure 1). In 1855, Russia and Japan signed the Treaty of Shimoda, which divided the northern and southern islands between them. One island, named Sakhalin, contained Russian, Japanese, and Ainu inhabitants, and so was omitted from the Treaty. Hostilities increased among the inhabitants of Sakhalin, and in 1875 Russia and Japan signed the Treaty of St. Petersburg, which gave sovereignty of Sakhalin to Russia and sovereignty of all the remaining Kuril Islands to Japan. As Bobic notes, the interest in the Kuril Islands is largely strategic. The islands have held strategic significance especially for Soviet Russia, which sought to station submarines in the area. Their economic value, however, is meager, as there are few petroleum or mineral deposits. There is some possibility of oil and gas reserves, but the amount is unknown. Finally, the islands have symbolic significance, insofar as they have been the site of important violent struggles between Russia and Japan. As Bobic reports, “the symbolic value of the islands matters the most to the local Russian residents, who believe that this was the land won with the blood of Russian soldiers.” [10] Early in the twentieth century, the dispute escalated into full conflict between Russia and Japan, which was eventually resolved through ratification of a peace treaty through the mediation of the United States. After the October Revolution of 1917, Russian forces again clashed violently with Japanese in the region, but this eventually led to another agreement, the Peking Convention of 1925. Following the conclusion of World War II, Stalin expressed his desire to seize the Kuril and Sakhalin Islands from Japan, and did so with Roosevelt’s blessing in the Yalta Agreement of 1945 (see Figure 2). Japanese-Russian relations were “normalized” in 1955, but the dispute over the islands remained. Late in the 1970s, the Soviet Union stationed troops on some of the islands, and a few years later Japan sent Prime Minister Suzuki to visit the southern islands in the archipelago, and designated a “Northern Territories Day,” which only served to escalate tensions. At last, starting in 1990, Yeltsin in Russia moved toward a proper resolution of the dispute, and met with Japanese officials in 1993. Eventually, this led to an agreement of mutual use of fisheries in the region and of visafree travel for Japanese to the area. Most recently Putin, however, has stalled further talks on resolution of the dispute. (See Figure 3 in the Supplementary Data provided at http://ontology.buffalo.edu/14/territorial-disputes/.)
Our ontological approach is capable of representing this complex situation, spanning over a century of tension, conflict, treaties, and shifting borders. We do this by first considering the timeline formed by a sequence of acts of treaty signing. Each such act has the output of a signed treaty, and it also occurs at a specific temporal interval, which can be timestamped. Thus, there is an act of treaty signing with output the Treaty of St. Petersburg. This act has two participants – the governments of Japan and Russia – and occurs at a temporal interval that is designated ‘1875’. Such temporal intervals are proper parts of the temporal interval occupied by the territorial dispute taken as a whole. Two discrete increases in conflict, resulting in armed conflict, occur on their own temporal intervals, designated by dates, and each overlaps with parts of the temporal interval of the dispute. (See Figure 2 figures in accompanying material.)
In addition, the symbolic importance of the islands rests on an instance of the disposition type we have labeled nationalism, and this symbolic importance serves as an ideological motive for the dispute. The objectives of the governments, on the other hand, turned on strategic naval advantage. Both motives and objectives are distinct from the explicit arguments put forward by participants in the dispute at different times through the century. Acts with arguments as outputs can be represented in their turn as occurring on specific temporal intervals that are designated by particular dates.
CONCLUSION
We have surveyed the ways in which our ontological approach can capture the features of a territorial dispute within the framework of the Basic Formal Ontology by appealing to the I2WD ontology suite. In particular, the ontology we propose offers the ability to capture the peculiar character of disputes and associated conflicts, it has a strategy to deal with both false and disputed information, and with the various kinds of arguments, motives, and objectives at work within them. In addition to terms representing entities such as claims, arguments, territories, and roles, the ontology must specify also the relationships among the diverse elements involved, for example, the relationship between the content of a piece of propaganda and an objective, or between an argument and a claim, relations such as aboutness, support, ratified by, and so on. Some of these relations are illustrated in the Figures.
ACKNOWLEDGEMENTS
We wish to thank John Beverley for incisive comments on an early draft of this paper, and Ron Rudnicki and Yonatan Schreiber for fruitful discussion of these issues.
SUPPLEMENTARY DATA