Understanding Shifting Triadic Relationships in the Al-Qaeda/ISIS Faction Ecosystem

We propose and investigate 14 hypotheses linking changes in the intensity of relationships between two factions in the Al-Qaeda/ISIS (AQ/ISIS) (ecosystem to future changes in other relationships involving one of those two factions. Using a novel 28-year data set of relationships between factions of Al-Qaeda and the Islamic State (including their predecessor organizations) encoded as a time series of 267 signed weighted networks, we identify triangles of any set of connected factions $(u,v,z)$ and develop hypotheses capturing how the intensity of relationships between two factions $u,v$ changes when the relationship between either factions $z,u$ or factions $z,v$ changes. We investigate how the strength of these relationships changes in the short term (less than two years), medium term (two to four years), and long term (five to ten years). Surprisingly, we show that AQ/ISIS triangles that are not in balance (according to the balance theory) are unlikely to move toward balance as suggested by the balance theory, but those that are in balance will likely stay balanced (as suggested by the balance theory). These findings provide better insight into the inner workings and complex dynamics that shape the AQ/ISIS competition worldwide. They are also relevant to future developments in global jihadism, as new factions may emerge that seek to gain affiliative market share in the jihadist world at the expense of more established organizations.