There is one more thing worth mentioning here. There are discussions in the functional programming communities if models they design are anemic too. It is because in functional programming the usage of classes is not mandatory and sometimes not even possible. Even when it is possible, it could be not natural to use classes because functions and functional composition could solve many problems easier. I would argue that in case the behavior is still modeled and implemented as functions, such a model is not anemic. It might not be directly fitting to the original definition, but 2003 was the time of dominance for object-oriented programming languages, so it was natural to use the term "object model." However, the key here is in a combination of data and behavior and definitely when using rich type system combined with clearly defined functions that express the intent, such model is indeed not anemic.