Eckman, F. R. (2008). Typological markedness and second language phonology. Phonology and second language acquisition, 36, 95-115.
Eckman got a little bit philosophical when he started to defend typological markedness. His view is that since a higher-level explanation is not available, we may just stick to the lower-level explanation. This view is what Thomas Kuhn would call a “normal science”, a paradigm that has great explanatory and predicting power within its own framework, while making up ad hoc arguments to explain away anomalies.
It is true that markedness can predict errors in one’s interlanguage (e.g. final devocing, preference for less complicated structures, etc.). It is also true that interlanguage does reflect typological markedness (e.g. if one can produce /spl/, one will have no problem producing /pl/). There are, however, outliers to MDH and SCH, just as there are outliers to typological markedness. For example, French speakers prefer a more marked /ø/ over /ə/, when epenthesizing English clusters; Arabic speakers replace /p/ with /b/ in their English speech; some children acquire /tʃ/ before acquiring /t/, etc.. The problem with theories of markedness is not to show how much evidence supports their thesis, but how to account for all those anomalies.
According to Karl Popper’s classic view of falsification, one anomaly is enough to falsify the whole hypothesis, and ignoring the outliers and/or making up ad hoc explanations are, in Popper’s opinion, pseudoscience. The validity of MDH and SCH is thus greatly weakened every time an anomaly emerges.
There are, of courses, merits to typological markedness and Eckman’s two hypotheses, which do give us a description of our preferences for certain linguistic structures, but they do not tell us how these preferences are acquired. Does the preference for the unmarked reflect the aerodynamic nature of our speech organs, just like John Ohala proposed? Or was it determined by the perceptual system in our brain? I do think we should look for higher-level explanations, which account for not only language acquisition, but also other cognitive behaviors. Two theories that I think might have a higher explanatory power are Karl Friston’s free energy principle on action and perception and John Goldsmith’s application of information theory in phonology.