As dim as the prospects may be for scientists to try to bridge their differences, the U. Penn researchers believe there may be a path forward. Not only would such a move resolve the tensions experienced by these “human” scientists, but it would help advance the scientific enterprise. Indeed, the authors note that “most scholars agree that the goal of science is to build knowledge about empirical reality and pursue truth by testing predictions and explanations against data”.

The path forward, Clark et al. proposed, involves adopting eight simple and non-so-simple steps to create teams of “adversarial collaborators.” Although designed specifically to apply to scientists embroiled in longstanding rivalries, their general principles are easily translated into practical tips anyone can adapt to their situations.

1. Consider the temperaments of potential adversaries

Some scholars may be able to participate in adversarial collaborations more successfully than others (e.g., successful adversarial collaboration may be associated with higher intellectual humility, open- mindedness, and agreeableness, and with lower dogmatism, neuroticism, narcissism, and ideological extremism). For many scientific disputes, different “sides” are supported by numerous scholars, and so it may be useful to select an adversary among them who seems capable of carrying out an adversarial collaboration successfully.

2. Involve a trusted, neutral third-party colleague to be a moderator.

The moderator should be mutually agreed upon by all adversaries and will coordinate the effort, referee disagreements, and collect and analyze the data and write up the results. The data should remain under the control of the moderator throughout the project. At the outset, the adversaries and the moderator should agree that the moderator will pursue publication even if one or more adversaries refuses to cooperate and drops out. (

3. An initial discussion should identify a clearly defined disagreement.

Both sides should be able to articulate their own perspective in concrete terms as well as the strongest version of their adversary’s perspective and the disagreement in terms all parties agree with. This discussion should leave all parties feeling understood, not caricatured. The moderator should take notes of all discussions; this allows for records that remind adversaries of their earlier statements and commitments.

4. Agree on the details of an initial study designed to subject the opposing claims to an informative empirical test.

The participants should seek to identify results that would change their mind, at least to some extent, and should explicitly anticipate their interpretations of outcomes that would be inconsistent with their theoretical expectations.

5. Strive for achievable, incremental progress.

Accept in advance that the initial study will be inconclusive. Allow each side to propose additional experiments to exploit the fount of hindsight wisdom that commonly becomes available when disliked results are obtained. Additional studies should be planned jointly, with the moderator resolving disagreements as they occur.

6. Be flexible with collaborators.

There is rarely one way to answer a question, so if there is resistance to one approach, simply move on to a new one. If one study goes awry (i.e., one or more collaborators are not convinced by the findings), figure out why and fix the ambiguities for the next study.

7. Take advantage of preregistration.

Preregistering an adversarial collaboration can help lock both scholars into a research plan, which will minimize scholars’ ability to renege if unfavorable results are found.

8. If significant disagreements remain after all data are collected, write individual discussion sections.

The length of these discussions should be determined in advance and monitored by the moderator.

If we can’t move forward alongside our enemies, then we should at least be able to move on separately from them. As the The U. Penn researchers note:

“These guidelines anticipate ways in which adversarial collaborations can fail and aim to pre-empt them. Although some adversarial collaborations proceed harmoniously, in which case these precautions are overkill, we think it wise to prepare for the worst and hope for the best. Disputes with high symbolic or policy stakes can quickly become contentious.”