Thank you for your comments. Indeed, the way in which I introduced charges using thermodynamics is more suitable for students that have already been exposed to some thermodynamics (moreover, I illustrated the method using a “language” that can be easily understood by expert people that can then adapt it to their classes).
The case you mention is quite frequent, indeed, and it is not peculiar to charges. Physics textbooks (especially those at low level) tend to give “recipes” or “prescriptions” to students and that makes some topic difficult to understand (and, in few cases, even difficult to believe if compared to everyday experience).
That’s why I try to adopt a very different method: I usually make experimental observation before introducing a topic, then I introduce the topic as a formalisation of our observations. For example, concerning the two charges it is easy to show that you can change the “electric state” of some object by rubbing it (use, e.g., a paper towel and an inflatable balloon). Before rubbing it, the balloon can’t attract paper, while it can after. That is interpreted saying that the object changed its “electric state”. We say that it acquired a “property” that we call “charge”. Two balloons, rubbed, repel each other, so we conclude that two “charges” repel each other. Rubbing other insulating materials (e.g. glass), you experimentally observe the same behaviour (i.e. the glass changes its state becoming electrically charged), however, glass and plastic attract each other, so the state of the latter must be different. We only observe two different states that we call conventionally + and -. We could call them a and b or 1 and 0, but the usual way of classifying the charges has the advantage of automatically determine the direction of the force that manifest between them.
What physics textbook should teach, in my opinion, is how physics laws descend from observations. Physics laws are irrelevant for most people: the method used to derive them is not.