Given that the majority of the tone patters are still ambiguous due to the limited speaker pool and recording quality, all instances in the dictionary should reflect that uncertainty.
Thus @cert for all <pron> should be marked as "low" by default and those for which the tones are known for sure should be changed to "high".
Given that the majority of the tone patters are still ambiguous due to the limited speaker pool and recording quality, all instances in the dictionary should reflect that uncertainty.
Thus
@certfor all<pron>should be marked as "low" by default and those for which the tones are known for sure should be changed to "high".