However, as in the external reality there are a huge number of facts difficult to correlate, the method to find the fundamental symbolic model is guessing. Once the model built, this will order in a univoque manner all the primary facts. Moreover, it will make predictions that will lead to new discoveries or confirmations of itself. As we have shown in the general theory, we reflect sections of the external reality in models. The models make predictions. If the predictions are good, we will use the model a next time too, as it proved to be useful. Now we have the normal answer to a fundamental question asked for long time: 'why do the laws of nature exist?' or 'Why the world has an order?' As it results from MDT, we reflect the external reality based on symbolic models. These symbolic models need to be logical in order to be stable. If a symbolic model associated to external reality will not reflect it correctly, we will build a new model. Example: The external reality can change due to the movement of some objects. Then we will build a symbolic model containing the term 'velocity'. This symbolic model will make good predictions provided the objects move at a constant velocity. If the velocity is not constant, the model will not reflect correctly the external reality. Then we will build a new symbolic model, introducing a new element called 'acceleration'. This model will make correct prediction for the objects that move at variable velocities as well. Thus, by building of adequate models, the external reality is reflected by stable models. It is very easy to confuse the external reality with its 'image' generated by a stable symbolic model. As we know, we have no direct acces to external reality. We perceive it based on some associated models. Thus, as a conclusion, the impresion that nature is a structure based on stable laws and order comes from the fact that we reflect the external reality based on logical and stable symbolic models. ETA 4: General Communication Language (GCL), dictionary GCL is the first symbolic model generated by the human brain. At the beginning, only its spoken form existed, later it appeared as a written language too. As the purpose of the construction of models is to predict the evolution of external reality, GCL was always associated directly or indirectly with external reality. GCL is a very special symbolic model. It is used both for general communication and for building other more precise symbolic models. Examples of developed languages included in GCL: diplomatic language, juridic language, logical and mathematical language, languages based on gestures and signs, computer languages. GCL can be used to build symbolic models that are associated to external reality, e.g. the positive sciences.