Actually symptoms being unobservable is not only not unique to psychology (the distinction is well known in medicine), it's a core challenge within the practice

Actually symptoms being unobservable is not only not unique to psychology (the distinction is well known in medicine), it's a core challenge within the practice of medicine and one of its historic tensions. Yeah, a patient might be complaining of chest pain, but the reason for the complaint is going to be unknown to you until you start the process of diagnosis. Chest pain is a symptom of many conditions, and you can't even trust that there's an observable cause, you can only believe what the patient tells you. This was obviously even worse until modern developments of things like MRIs and X-rays. Compare that to signs, where you can see that a patient has say, a lesion or is clearly coughing.