To avoid perpetuating an armed conflict that has already lasted over 50 years, the Colombian government is talking with its political opposition, led by former President Álvaro Uribe of the right-wing Democratic Center Party, as well as FARC leaders, to try to modify the previous peace agreement to one that all three sides would support and that could, hopefully, gain broad public legitimacy. With that in mind, government negotiators renewed talks with members of the FARC this weekend in Havana, bringing with them proposals from the various parties who voted “No” in this month’s referendum.

Lessons from previous conflicts can be helpful, but Colombia is also charting new territory as a country with strong legal and democratic credentials and a commitment to new international humanitarian and human rights obligations. Add to this the fact that Colombia took the unusual step of consulting with voters in a plebiscite, seeking to gain legitimacy for the accords. Notably, the dissenting voters were concentrated in parts of the country relatively unaffected by the long armed insurgency, while voters in conflict zones appeared more eager to put an immediate end to the conflict.

The primary points of contention among voters and negotiators involve the commonly perceived trade-off between peace (with amnesty) and justice (with prosecutions), as well as the extent to which FARC leaders can participate in Colombian political life. In the past, the majority of post-conflict and transition agreements in authoritarian or post-war states granted broad amnesties for human rights abuses under the notion that “forgive and forget” was the best path forward to avoid backlash and a possible return to violence.

Exceptions to this rule happened when leaders of the former regime were defeated and/or eliminated from public life (Germany and Japan after World War II, Argentina, and Greece); a truth commission was established to provide full “confession-for-amnesty” solutions (South Africa and now Colombia); or the UN Security Council passed legally binding resolutions mandating prosecutions by supranational or hybrid national-international courts (International Tribunal for the former Yugoslavia, International Tribunal for Rwanda, Special Court for Sierra Leone, War Crimes Chamber of the State Court of Bosnia and Herzogovina, equivalents in Croatia and Serbia, and the newly established Specialist Chambers and Specialist Prosecutor’s Office for Kosovo).

Most other transitional or post-conflict regimes have not prosecuted anyone in the short run or at all, including the three dozen countries that have implemented truth commissions or other investigatory bodies to analyze past crimes. A few domestic courts have attempted prosecutions and failed (for example, the Special Court for East Timor and the Special Court for Lebanon).

Since the International Criminal Court (ICC) and its 1998 Rome Statute came into legally binding force in 2002, states parties, including Colombia, have been legally required to either investigate and prosecute those most responsible for committing crimes against humanity, genocide, and/or war crimes, or, if unwilling or unable to do so, have the ICC prosecute them. Thus, broad amnesty for human rights crimes that had also helped induce either armed groups to lay down their weapons or authoritarian governments to transition to democracy are no longer acceptable under international law.

Many analysts of earlier conflicts concluded that, when those most responsible for mass atrocities are not prosecuted and punished in the present, signals of tolerance for impunity are sent out for the future. Given the variety of contexts and factors affecting transitional justice outcomes, no clear track record offers a “one-size-fits-all” strategy among the menu of post-conflict accountability mechanisms. This menu ranges from the most audacious step of prosecuting authors of atrocities and even using torture, to other, non-punitive forms of accountability, such as truth commissions, purges, vetting of human rights violators, and filing lawsuits for reparations and public confessions.

Some countries, like Chile and Brazil, have tried using truth commissions to bring information to victims’ families without immediately punishing violators. Argentina went further and prosecuted six leading commanders in the mid-1980s. That initial stand against impunity was more than the new democracy could withstand, however, and led to three coup attempts and several mutinies between 1987 and 1988, and even the subsequent pardoning of those convictions by the next president. Fifteen years later, though, the country began convicting thousands of abusers, by far the most prosecutions of any country ever. Present-day Colombia, it should be noted, is a more established democracy.

It remains to be seen whether countries will have the political will to enforce the principle that crimes against humanity have no statute of limitations and choose to prosecute violent human rights abusers, after a  new generation comes to power in politics, as well as in militaries. The process has always been complicated and controversial. This has been perhaps nowhere more true than in Spain, where the leading investigative judge, who initiated the prosecution of human rights abuses from the Franco regime, Baltasar Garzón, was suspended from the bench after complaining that human rights were being repressed in Spain in recent years by forces sympathetic to the Franco dictatorship. Likewise, Chile began prosecutions more than a decade after the fall of the Pinochet government, but only after an outsider, the Spanish judge Garzón, requested Pinochet’s extradition for trial in Spain, and opened up the debate in Chile.  El Salvador’s Supreme Court, just this year, struck down their 1990s amnesty laws that had been negotiated through United Nations mediation.

The Colombian peace negotiators, and supporters of the September 26, 2016 agreement, argued that both peace and justice could be achieved by setting conditions for reduced or alternative sentencing. In the original agreement, combatants on both sides—FARC and state security forces —who are guilty of grave human rights crimes yet decide to confess the truth, pay reparations, and acknowledge responsibility would be eligible for a reduced sentence of no more than eight years. The guerrillas would not be sent to traditional prisons, but would have restricted liberty in designated geographic areas where they would be required to perform community labor to repair damage done to their victims. Those who do not comply with these conditions would face a special court, and prison sentences of up to 20 years.

The FARC would have six seats in Congress with a voice, but no vote, until the next elections are held in 2018. From 2018 to 2026, they would have a guaranteed minimum of five seats in the House and in the Senate, but could compete for other legislative posts as well.

Several prominent politicians who opposed the agreement have made proposals to modify it, and these are less strident than many observers expected. Former President Álvaro Uribe’s political party proposed that both the military and the FARC be investigated and prosecuted under the regular justice system, rather than the agreement’s proposed Special Tribunal and Jurisdiction for Peace. The original agreement had also proposed creating mixed special tribunals, featuring mostly national, but also a small number of foreign judges, as has previously been done in Lebanon, Sierra Leone, Cambodia, and Bosnia. The formula to create a special transitional justice mechanism had, as a further goal, to acknowledge the role of victims as individual rights holders.

The new proposals would aim to make the agreement more palatable to those who opposed it by making the punishment of grave human rights crimes more stringent through a requirement of confinement to labor farms. In potential deal-breakers for the FARC, the proposals would prohibit those guilty of war crimes and of crimes against humanity from ever participating in politics, remove drug traffickers from amnesty eligibility, and limit the land reform provisions to protect the property rights of large landowners (deep land inequality was one of the origins of the conflict in the 1960s).

Another factor that contributes to sustainable peace is international involvement. Such support could involve multilateral peacekeeping troops and police enforcing demobilization and disarmament processes in order to deter future violence. Civil monitors from abroad may also assist or advise in the prosecution process—which must be above reproach to avoid becoming a source of additional conflict. Multilateral peacebuilding, which has had a very uneven record worldwide since its introduction by then UN Secretary General Boutros Boutros-Ghali in 1993, also requires significant foreign aid and technical assistance, yet must be without foreign meddling. Such assistance can help local institutions and groups achieve security or justice reforms on their own, as well as compensate victims, fund truth commissions, retrain and reintegrate ex-combatants into productive civilian lives, and develop impoverished areas of the country—often the conflict zones, long-neglected by, or inaccessible to, state services.

As a more advanced democracy, Colombia has more potential to achieve real reforms than did the uneven or failing attempts to build peace in Bosnia and Kosovo, for example. In those cases, foreign troops have only maintained a negative peace, without actually integrating former enemies into political and civil society.

Despite opposition from inside Colombia, international support was strong for the original agreement, and the United Nations had already set up a monitoring force to demobilize and disarm the FARC guerrillas as they eventually gather inside special zones. This has now been put on hold and full funding has yet to be determined to pay for the high costs that ending the conflict will necessarily incur. A United States “Peace Colombia” plan has still not been approved by a Congressional Conference Committee and is also subject to the U.S. electoral outcome. Should it be approved, it would mean the provision of $450 million more than the current annual U.S. allocation of $320 million that goes to peacebuilding programs in Colombia. And several other countries also plan to contribute to a de-mining fund.

But the Colombian government will need to raise its own funds as well to meet the ambitious rural development goals, as well as other aspects of the peace agreement. President Santos had been holding off on introducing a new fiscal reform bill to raise these revenues until after the October 2 plebiscite, but has now had to move forward with this controversial package, introducing it in Congress last week.

And so Colombia remains in limbo, but, at the same time, in need of an urgent resolution to existing disputes over the accord. Fortunately, the government and the FARC extended the bilateral ceasefire through the end of 2016, and all sides now appear willing to consider some modifications to the original agreement. Separately, the government also announced the beginning of talks with the one remaining, smaller insurgency group—the ELN (or National Liberation Army).

Unfortunately, however, many risks remain that could threaten the success of renewed negotiations. These include the politicization of the process, as the 2018 presidential election season approaches in Colombia—in which the opposition could calculate that delaying peace would be a worthwhile political gain. Even more immediate is the risk posed by the 7,000 FARC guerrillas and additional thousands of civilian supporters in FARC-dominated areas. They fear for their own security and will lack proper means of financial support if they continue to suspend drug trafficking and other illegal revenue-generating activities, as promised during the ceasefire.

Finally, the widespread existence of other criminal gangs and drug cartels in Colombia are poised to fill vacuums as the FARC recedes. FARC troops facing a very uncertain future may lose confidence in their own leaders, as well as in the government, leading them to join these criminal elements. What this all means, in the end, is that time is truly of the essence if negotiators are to reach a formula that can enjoy broad legitimacy and hope to bring lasting peace to Colombia.