Present human languages display slightly asymmetric log-normal (Gauss) distribution for size [1-3], whereas present cities follow power law (Pareto-Zipf law). Our model considers the competition between languages and that between cities in terms of growing (multiplicative noise process) and fragmentation ; where, relevant parameters are (naturally) different for languages and cities. We consider lifetime distribution for old and living languages and that for old and living cities. We study also the effect of random elimination (punctuation) within time evolution of languages and cities. Finally, we assume decreasing exponential distribution for cities over size with independent random amplitude and random (negative) exponent; and show that, this gives the Pareto-Zipf law.