I’ve found that LDM (Logical Data Modelling) is primarily an issue of semantics, concerned with the perfection of the semantic representation of a database more than its structural or relational aspects. To some extent, this is quite frustrating, as it reminds me of the French Ministry of Language, whose duty is to maintain the beauty and purity of French. Does it matter if email in French becomes courriel instead of an Anglicism? Not particularly: they both encode the same information. But, according to the mininistry, only courriel may be used. In the same way, when creating a logical data model, there are a strict set of arbitrary guidelines to shape the nomenclature of your data model. And, while they set standards of clarity and uniformity, they also prevent a high barrier of entry into the world of data modelling without producing better end results–better databases.

My unanswered question: why does a logical data model have to follow strict semantic rules?