This is the last paper in a Comments and Controversies series dedicated to "The identification of interacting networks in the brain using fMRI: Model selection, causality and deconvolution". We argue that discovering effective connectivity depends critically on the adoption of a state-space model with biophysically justified observation and state equations, that will however have to be augmented with proper specification of priors, and a check for model Identifiability. We analyze similarities and differences of Dynamic Causal Modeling, Granger Causal Modeling and other approaches. We establish links with past and current statistical causal modeling. in terms of Bayes-net, Wiener-Akaike-Granger-Schweder influence measures, and their synthesis.
We show that some challenges posed to the general approach may be open to solutions, and speculate on promising areas for future development.

（脳科学研究戦略推進プログラム　課題D　研究者向け）