According to Good (1988) and Gulati (1995), trust is the product of the repeated meeting of actors. This claim jeopardizes (or perhaps merely complicates) the distinction between bond and bridge. After how many meetings does a bridge become a bond? And does the transformation entail loss, increase, or maintenance of the bridge’s resources? That a connection cannot be definable as simultaneously bridge and bond seems improbable, if trust – the elemental contribution of social capital – is the product of repetitious interaction alone, regardless of interaction quality.
The bridge and bond dichotomy fails to place connections that begin as bonds but evolve into or occasionally act as bridges, and vice versa. The bridge/bond dichotomy enforces stasis (a highly uncommon observation in sociological study) – neither bridge nor bond can change states. Defining the boundaries of networks is also problematic, and would likely become increasingly so as tie counts rise and relationships evolve.
Jones and Conway’s examination of British entrepreneur James Dyson’s network (2004) reveals incompliant instances. Although this study is quality-sensitive and of questionable extensibility (as all single case studies are), it demonstrates that bonds are capable of acting as bridge-like intermediaries, granting access to instrumental resources and even upward-reaching social cleavages. This study puts Putnam’s absolutism to test. Exceptions appear both possible and likely.
The findings of Newell et al (2004), by suggesting strong ties are fundamental to knowledge acquisition, also challenge the bridge/bond resource sets. So too with Rodan and Galunic (2004), whose findings show that a sparse collection of weak ties provides enhanced value when network heterogeneity is available.
The egocentric protocol of network depiction (as shown in A.5.) might be susceptible to psychocentric bias. Connections are defined as bridges or bonds according to their relationship with the ego. Sociocentric possibilities are by inference alone, i.e. a bridge could possibly link to another bridge, but the egocentric protocol does not give scope for inter-network interaction. Moreover, the appointment of an individual as the ego from whom bridges and bonds radiate contravenes schools of thought that posit innovation to be an intrinsically network-derived phenomenon (e.g. Rothwell and Zegveld, 1985).