When we define the term 'government' in syntax, we should exclude the interfering nodes which asymmetically c-command node B.Why can't we allow asymmetrical c-command while we accept symmetrical c-command? In other words, why should we allow the interfering nodes which symmetrically c-command B? Please let me know, please. :)

  • Looks like homework ;) Unless you show some effort of your own and tell us what you have come up with yourself, people are unlikely to put in any effort themselves.
    – robert
    Oct 12, 2014 at 13:25
  • 1
    The actual reason is that somebody said so once in their MIT dissertation and it became part of canonical wisdom. But nobody remembers who said it, or when, or why, nor cares that the arguments no longer make sense in modern theories. It's just the Way Things Are, that's all. Resistance is futile; you will accept symmetric c-command.
    – jlawler
    Oct 12, 2014 at 15:21
  • I do not understand the question, but I support jlawler's comment. Any syntax course that is building on c-command in our modern times is ignoring the fact that c-command has been demonstrated as much less than insightful in numerous ways by many, concerning numerous phenomena of syntax (binding, scope, etc.). The reason c-command survives can only be explained in terms of institutional authority and the laziness of many living syntacticians. Oct 13, 2014 at 4:01

1 Answer 1


I have troubles parsing the question unambiguously. My answer assumes the question means "Why are symmetrical c-commanding nodes not barriers to government whereas asymmetrically c-commanding nodes are (potentially) barriers to government?" but I encourage you to rewrite the question with more details in case I misunderstood.

Under this assumption, the historically accurate answer was that this was a pure stipulation based on the fact that such an ad hoc fix gets the case assignment and binding properties right. The main problem with allowing symmetrical c-commanding nodes as barriers is the inherent ambiguity this involves: since binary trees have no ordering (so that their common drawing is deceptive), it would be impossible to define which node is a barrier for the other in the generic situation were symmetric c-commanding nodes to be allowed.

However, such stipulations should always be regarded with great skepticism and accordingly a good deal of works in that branch of syntax in the last 30 years has been devoted to eliminate them (the impetus behind the so-called minimalist program). Within the framework of minimalist syntax, the answer to your question follows from the data structure of the syntactic objects constructed by Merge, as I encourage you to work out for yourself (or ask about).

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.