Transduction from an explicit unambiguous internal semantic representation to a word sequence is much less problematic than reconstruction of an explicit semantic representation from an ambiguous word sequence
Transduction(as used in linguistics and cognitive science)
The process of converting information from one form into another—in this case, taking a clear thought in your mind and turning it into spoken or written words.
Unambiguous(as used in this philosophical statement)
Having only one possible meaning; the opposite of unclear or confusing.
Word sequence(as used in linguistics)
A string of words arranged in order, like the sentence you're reading right now.
ambiguous(in logical analysis)
Capable of being understood in more than one way; having multiple possible meanings.
Despite this inverse relationship, language generation has traditionally received less attention from computational linguists than language understanding, because if the content to be verbalized is available in an unambiguous, formal representation, standard output templates can often be used for generation, at least in sufficiently narrow domains. Even for unrestricted domains, the transduction from an explicit, unambiguous internal semantic representation to a word sequence is much less proble