A model, which belongs to PSM, cannot be changed regardless of the information received from external reality (the PSM contains only invariant models). In special conditions, e.g. when a big danger exists for the being (as detected by PSM), it is possible that a new model enters the PSM. Basically speaking, any model can enter the PSM. For a normal brain, the PSM must contain only "standard models" (see the general theory and ETAs) because, once a model is in PSM, it cannot be changed regardless of the information received from external reality. Even more, any information from external reality can be accepted only if it can be accepted by PSM. Example: Let's suppose that in the PSM of a person there is a non-standard model which considers that the frogs are very dangerous. Regardless of the information received from external reality, that person will be horrified when frogs are around. The content of the PSM is very hard to be known because the PSM is activated only when there is no normal model to understand the external reality. The PSM is an image model and it will remain so forever. EXAMPLES, TESTS AND APPLICATIONS (ETA) ASSOCIATED TO THE MDT THEORY These ETAs are intergrant parts of MDT and show how it works in some specific cases. The order of occurence of the subjects is random. MDT tries to keep its generality as much as possible, independent of the technological implementation of different brains. ETA 1: The Model The model is a collection of elements and relations between the elements. There are two types of models: image models (or analogic models) and symbolic models. The elements and relationships are given explicitly for the symbolic models, and implicitly for the image models. Image models (analogic) can't be given in an explicit manner. They are given as they are, as a whole. This is an intrinsic property of the image models. To give a model in an explcit manner means to describe the elements and the relationship between the elements, but this takes us outside the analogic model. That means to translate the image model into a symbolic model (we need to use words to describe the image model). Even if the translated model is associated to the image model, it is a different model. Example: given an image model of an airplane, its elements are the main body, the wings etc. One of the wings could break in two, so it is made of two pieces. Actually, it contains an infinity of elements, as it could break in any way. In any real situation, it is by far easier to build an image model, than explain what had been built. This is why we say that an image model is just given as it is, and not defined explicitly. Anytime we refer to an image model, we need to take into account this fundamental issue. Application 1: Image models in poetry and painting