Let's start by explaining a use case and the confusion it gives due to poor nomenclature in the crypto-world. The case of me giving a so-called private-key to someone, so I can send the person an encrypted message with the pared public-key, so only the receiver can open it. In this use case, in fact, I make a private-key public and assume the public-key to be private. This use case makes clear that the naming of private-key and public-key are a poor semantic choice. It is a poor choice because depending on the use case, the private-key can be public and the other way around. We better call the private-key just the encryption-key, and the public-key an encryption-vault. Let's then revisit the example use case now with the new naming. The case of me giving a so-called encryption-key to someone, so I can send the person an encrypted message in an encryption-vault, so only the receiver can open it. We have 4 clearly define objects now: public-encryption-key private-encryption-key