This post gives an account of the processes that are involved in the learning of anxiety. More specifically, it will consider the theory of classical conditioning.
Classical conditioning was originally illustrated by Pavlov in his dog experiments (read more), and the experiment showed how a specific stimulus (food) could evoke an immediate unconditioned response (UR), such as the salivation in dogs.
Then Pavlov illustrated how this response could be conditioned, which implies that the salivation could become activated without the presence of the specific stimulus (food).
The conditioning could be evoked by the sound of a metronome, and whenever the dog was fed, the sound was played. Soon the dog associated the sound with the presence of food.
Therefore, the sound of a metronome was enough for the dog to produce salivation because it had associated the sound with the presence of food, and this response is called a conditioned response (CR).
You might think, what does this have to do with anxiety? Well, in the same way a dog becomes conditioned, the same way anxiety or a phobia can become a conditioned response. It can therefore originate without the presence of a direct threat (a noxious stimulus).
According to the learning paradigm, the anxiety is caused by a learning process. Anxiety may originate from an episode, where the individual has learned that such an episode is (life) threatening, and that similar episodes are threatening as well.
For example, being arachnophobic (i.e., the fear of spiders) can be due to an episode where a given spider has been perceived as a direct threat, but soon after, the threat becomes conditioned, and then the phobia is provoked by the mere talk of spiders. It is therefore not the spider itself, but rather the mental representation of the spider that causes the anxiety.
Theoretically, there are no limits to how many associations an individual can make, and the anxiety will just get more pronounced with each association made. This is certainly an irrational response, but it is a way to navigate through possible threats in the environment. So, anxiety makes one careful, but also too careful.
It even seems that some phobias are innate, or at least readily triggered. This is probably the case with arachnophobia (see this link). It may simply be an evolutionary advantage to be afraid of spiders, just like it is to be afraid of predators or heights.
However, if you do not live in a place where deadly or dangerous spiders exist, the fear of spiders is irrational, and you are then being too careful as the threat is nonexistent. I hope you get my point.
When a person has learned a conditioned response, it is very difficult to extinct the response, and indeed an older study by Eysenck (1968) shows how anxiety maintains for long period of time without the influence of a direct threat. When one is afraid of something, one tries to avoid threatening situations, in which the anxiety is being provoked.
This may lead to avoidance behaviors, which in fact feeds the anxiety even more. This is because, if we persistently avoid the situations that provoke the anxieties, then the anxieties are never confronted and desensitized.
Also, we will never learn that the anxiety is disproportional, and we will therefore remain in a state of anxiety. The anxiety will add up and overwhelm our nervous system sooner or later.