5. Discussion
5.1. Discussion of research findings Hypothesis 1a predicted that SL subsidiaries will prefer to engage in internal knowledge sourcing. However, our findings indicate the opposite (i.e. a preference for external home knowledge sourcing). This might be explained by the fact that, over time, this type of R&D subsidiary is likely to develop its own capabilities and knowledge and evolve to become a different type of unit (Birkinshaw & Hood, 1998), which potentially might weaken its ties with the intra-MNE network. In other words, over time, the organizational and knowledge processing structures of SL subsidiaries are likely to change, resulting in a greater focus on external rather than internal knowledge sourcing. Our findings do not support Hypothesis 1b about the relationship between LIL subsidiaries and knowledge sourcing. We know that LILs are mandated to coordinate closely with various other functions in their local environment in order to develop or enhance products according to local needs and scope with the result that this type of subsidiary functioning falls within SLs and IILs subsidiaries. A possible explanation for this lack of support for Hypothesis 1b is that LIL subsidiaries undertake a wide range of research or production/adaptation work when compared to the narrower mandate of the other two types of R&D subsidiaries. Hence, this diversity might have affected the survey responses and produced a large variation in the responses to the questions on knowledge sourcing.