On the ‘Definability of Definable’ Problem of Alfred Tarski