Emotions(like fear,anger,sadness,happiness etc.) are the fundamental features of human behavior and governs his/her mental health. The subtlety of emotional fluctuations can be examined through perturbation in conversations or speech. Analysis of emotional state of a person from acoustical features of speech signal leads to discovery of vital cues determining his or her mental health. Hence, it's an important field of research in the area of Human Computer Interaction(HCI). In a recent work we have shown that how the contrast in Hurst-Exponent calculated from the non-stationary and nonlinear aspects of "angry" and "sad" speech(spoken in English language) recordings in the Toronto-Emotional-Speech-Set(TESS) can be used for early detection and diagnosis of Alzheimer's Disease. In this work we have extended the work and extracted Hurst-exponent for the speech-signals of similar emotions but spoken in German language. It has been observed that the Hurst-exponent efficiently segregates the contrasting emotions of "anger" and "sadness" in the speech spoken in German language, in similar fashion it has been doing for English speech. Hence it can be concluded that the Hurst-exponent can differentiate among speech spoken out of different emotions in language-independent manner. We propose algorithm for a language-independent application for early non-invasive detection of various severe neurocognitive-disorders like Alzheimer's Disease, MND(motor-neuron-disorder), ASD(autism-spectrum-disorder), depression, suicidal-tendency etc. which is not possible with the state of the art medical science.