In NLU, you ignore intenSion at your peril

I have purposely wrote “intenSion” in the title with an uppercase ‘S’, so that it is clear I did not make a mistake (where it might be assumed that the correct word is “intention” — as in “intent”). So, I did not make a mistake, and my intention is “intension”!

It is one thing to be annoyed by MS Word that always considers ‘intension’ a spelling mistake — until I add it to the dictionary — but the fact that this is almost an unknown concept by most researchers and engineers in NLU) is puzzling. Indeed, while the notion of ‘intension’ is rampant in computer science, and especially in AI/NLU, I have always been amazed by how many researchers/engineers in AI never heard of ‘intension’ since you simply cannot build a system that understands natural language(s) without accounting for intension. You just cannot! So said Immanuel Kant and Gottlob Frege, and Rudolph Carnap, and Richard Montague, and … and I believe them. Let’s see why.

So What is ‘Intension’?

Intension is usually contrasted with ‘extension’ and the two notions are at the heart of what concepts and objects are and what does it mean for two objects to be ‘equal’ and when they are…