There are two different ways of understanding the notion of ‘ontological commitment ’. A question about ‘what is said to be’ by a theory or ‘what a theory says there is’ deals with ‘explicit’ commitment ; a question about the ontological costs or preconditions of the truth of a theory concerns ‘implicit’ commitment. I defend a conception of ontological commitment as implicit commitment, and argue that existentially quantified idioms in natural language are implicitly, but not explicitly, committing. I use the (...) distinction between the two kinds of ontological commitment to diagnose a flaw in a widely used argument to the effect that existential quantification is not ontologically committing. (shrink)