As early as the beginning of the 17th century, the use of mechanical devices as a method of overcoming language barriers was suggested. This idea has developed considerably since then, with systems such as TAUM-METEO, Verbmobil and SYSTRAN currently in everyday use. This article analyses the current state of the art in Machine Translation, as well as defining the problems that make it so difficult The problems are then addressed individually with reference to the direct translation approach to machine translation. Each problem is then tested amongst four different systems, three of which are used commercially throughout the world. This knowledge is then analysed, to form a conclusion, stating how the direct translation methodology overcomes ambiguous problems.