I don’t see any mention in the docs of word stemming in DA. (Word stemming looks for different forms of a word–I don’t know about irregular forms.) If this were added it would definitely need to be able to be turned off.
DA doesn’t provide such a feature at the moment. All query terms are case insensitive and if "Ignore Umlauts" is activated, accents or umlauts are ignored (e.g. the characters e, é, è or ê are equal).