Antarctica’s treaties have a memory—and that may be their real strength

A treaty system can look weak because it has no world government behind it. Yet the Antarctic Treaty has managed something less dramatic and arguably more durable: states keep returning to old rules, revising some, retiring others and treating the whole body of decisions as connected. Casper van Elteren’s analysis of more than six decades of Antarctic decision-making finds that this maintenance work is not evenly spread. It clusters in specialized policy histories, especially the management of protected areas.

That is a strikingly practical picture of international cooperation. Agreements do not merely succeed or fail at the moment they are signed. They persist through the slower work of interpreting, updating and sometimes clearing away what was made before. In a place with no sovereign authority above participating states, that accumulated record can become part of how cooperation itself is sustained.

The study maps proposals, responses and links among decisions under the Antarctic Treaty. It distinguishes binding Measures from non-binding Resolutions, a distinction that matters because the two kinds of instrument do not appear to feed into one another in the same way. The result is not proof that the Antarctic system always achieves its goals; it is a map of how its rules are handled over time, and a way to ask sharper questions about cooperation elsewhere.

The Science

The Antarctic Treaty is often treated as a test case for a familiar complaint: international agreements have no police force, so states can ignore them. That complaint captures a real limitation. But it can obscure a second question: once states have agreed on rules, how do they keep those rules usable as conditions, institutions and priorities change?

Van Elteren focuses on the record of decision-making under the treaty system, not simply on whether states comply with individual obligations. The paper traces submitted proposals to recorded responses and examines how adopted instruments refer to, revise or otherwise act on earlier instruments. In other words, it treats decisions as a network across time rather than as isolated entries in a legal archive.

A network here means a set of items connected by relationships. The items are proposals and formal decisions; the links represent responses and references between them. This makes it possible to see whether decisions across different subjects share a common documentary fabric, whether certain policy areas repeatedly revisit their own past decisions, and whether different categories of instrument have distinct histories.

One central distinction is between Measures and Resolutions. Measures are the system’s binding instruments, while Resolutions are non-binding. The difference is not merely a label. If a binding rule tends to be built by modifying earlier binding rules, while a non-binding statement tends to have a different lineage, then the formal architecture of the system is doing real work in shaping how policy develops.

The paper also separates repeated revision from the administrative act of retiring old instruments. Periodic reviews can remove outdated decisions in bulk. Such a cleanup links decisions that might otherwise sit in separate policy histories, but it does not necessarily mean governments are coordinating substantive policy across those subjects. The distinction is important: an archive can become more connected because of housekeeping, without the underlying policies becoming more integrated.

The paper’s account is based on more than six decades of recorded decision-making. The abstract and supplied material do not provide the underlying counts, year-by-year series, model specifications or figure data. That limits what can responsibly be plotted or quantified here. The available evidence supports describing the reported patterns, but not inventing numerical estimates or reproducing charts that cannot be checked against the paper’s tables or figures.

What They Found

The first result is that shared references tie decisions together across the agenda. The Antarctic system is not simply a stack of independent agreements, each sealed off from the rest. Instruments point back to earlier instruments, creating documentary connections that cross subject areas and make the accumulated record part of the decision environment.

The second result is more uneven: recurring revision concentrates in specialized histories, above all the management of protected areas. That concentration exceeds what the paper says would be expected from the amount of activity in a subject, its popularity, or the age of the references involved. Put plainly, protected-area decisions are revised repeatedly not just because there are many of them, because states discuss the topic often, or because the underlying texts are old. The pattern points to a distinctive maintenance dynamic in that policy domain.

This matters because revision is often treated as a sign that a policy is unstable or that the original agreement failed. But repeated revision can also be what keeps a rule relevant. Protected areas are specific places governed by rules that may need adjustment as knowledge, conditions or management needs evolve. The study identifies a concentrated history of revision; it does not establish that every revision improved conservation outcomes. Still, it shows where the system’s attention to rule upkeep is most sustained.

A third pattern comes from periodic administrative review. These reviews retire outdated instruments in bulk, connecting otherwise separate histories. The connection is administrative rather than necessarily political: a review may gather old decisions from different areas into one process without coordinating the substantive policies in those areas. This is a useful warning against reading network connectivity as evidence of unified strategy.

The fourth finding concerns the legal layers. Binding Measures build overwhelmingly on earlier Measures, with little direct succession from non-binding Resolutions. That suggests a recognizable pathway within the binding layer: new binding rules tend to inherit from prior binding rules. Resolutions, meanwhile, have limited direct passage into the lineage of Measures, at least as represented in the decision links studied.

That separation does not mean non-binding texts are irrelevant. A Resolution could influence debates, expectations or later decisions without being cited as a direct predecessor. Nor does the pattern establish that Measures are always implemented. What it does show is that formal instruments do not all travel through the system in the same way. Legal status is associated with different patterns of documentary descent.

Together, the findings complicate the picture of treaties as either powerful commands or empty promises. The Antarctic case instead reveals a system maintained through consensus and an ongoing relationship with its own past. That relationship is visible in references, revision and retirement: the unglamorous mechanics that allow rules to accumulate without becoming a completely static archive.

Why This Changes Things

International cooperation is often measured at the endpoint: Did states meet a target? Did emissions fall? Was a protected place preserved? Those outcomes matter most, but they do not tell the whole story of how institutions function. A decision system can be poorly maintained even when its founding principles remain popular; conversely, a pattern of orderly revision can reveal institutional capacity even when outcomes remain contested.

The paper shifts attention toward this middle layer between agreement and result. A treaty is not only the promise made at a signing ceremony. It is also a sequence of later acts that interpret, amend, renew or retire the rules. Studying those acts can help researchers distinguish a system that is merely accumulating paperwork from one that has mechanisms for keeping its commitments legible and current.

Antarctica makes this question unusually clear. No state exercises sovereign authority over the continent in the ordinary way, and the treaty system relies on participating states working through shared institutions. There is no higher government that can simply rewrite the rules for them. Consensus is both a constraint and a governing method: decisions depend on states continuing to accept the process.

The results suggest that consensus need not mean every policy history moves in lockstep. Instead, the system can maintain specialized tracks, with particular areas receiving repeated attention, while administrative reviews connect separate tracks through common cleanup. This is a more textured picture than the idea that international governance must either centralize into a world authority or collapse into national self-interest.

It also offers a caution for people comparing institutions. A highly connected decision network is not automatically a coherent policy regime. Some links arise because old instruments are retired together, not because states have integrated the policies they cover. Likewise, a thin link between Resolutions and Measures does not prove that non-binding commitments have no influence. The network reveals formal traces; it cannot capture every informal discussion or political effect.

For the global commons—areas such as the high seas, outer space or the atmosphere, where no single state can govern the whole—this distinction could be consequential. These domains face changing scientific knowledge and competing national interests. They need agreements, but they also need ways to keep rules up to date and to distinguish substantive policy learning from administrative maintenance.

Van Elteren’s approach offers a possible comparative tool. If researchers map the lineage of decisions in different regimes, they can ask whether policy areas with frequent revision are also areas where conditions change quickly; whether binding rules tend to build on earlier binding rules elsewhere; or whether periodic reviews create cross-cutting connections without meaningful policy coordination. Those comparisons could clarify why some regimes adapt and others become brittle.

The paper’s contribution is therefore less a verdict on Antarctic diplomacy than a change of lens. Instead of asking only whether a treaty has teeth, it asks what the treaty system does with the teeth it already has: how it records them, sharpens them, and sometimes removes instruments that no longer fit. Rule maintenance is not a substitute for compliance or good outcomes. It is one of the conditions that may make enduring cooperation possible.

What’s Next

The most important next step is to connect the decision network to what happens outside the documents. Do revisions to protected-area instruments correspond to changes in monitoring, conservation practice or ecological conditions? Do states comply more consistently where a rule has a clear chain of prior decisions? Those questions would link institutional memory to measurable consequences.

A second challenge is to understand influence that leaves no citation. The finding that Measures rarely succeed directly from Resolutions concerns formal succession. Political ideas can travel through negotiations, expert advice or shared expectations without appearing as a reference in the final instrument. Future work could combine document networks with meeting records, interviews or evidence about implementation to capture those less visible pathways.

The authors’ account also invites comparisons beyond Antarctica. The same methods might be applied to governance of the high seas, climate agreements or other international regimes, while taking care that each has different legal structures and decision practices. A useful comparison would not simply rank systems by how many links their instruments contain. It would ask what the links mean: substantive policy development, routine review, or something in between.

There are limits to what can be concluded from the information available here. The supplied paper material includes its abstract and bibliographic details, but not the full article’s figures, numerical tables or methods. Accordingly, this digest does not report counts, percentages or plotted values, and it cannot responsibly reproduce three data charts or cite figure numbers from the paper. The findings above are those explicitly stated in the abstract; claims about causal effects or conservation success would go beyond them.

Even so, the central insight travels. Cooperation is not only the act of making a promise. It is the collective work of remembering what was promised, deciding what still applies and making changes that other participants can recognize as legitimate. In the Antarctic Treaty system, that work has a distinctive pattern: shared references across the agenda, concentrated revision in specialized histories, administrative cleanup across them, and a largely self-contained lineage for binding Measures.

That is not a guarantee of success. It is a reason to look more closely at institutional upkeep—and a reminder that the durability of global cooperation may depend as much on how rules are tended as on how boldly they are first written.