This paper attempts at studying the script variation causing the problem of text unification for Maithili language data collected from the various region of Bihar especially Maithili spoken area. Updating the devanagari script for developing the Maithili language corpora for Machine Translation (English-Maithili) and reverse. Many of the linguist & scholars (Maithili Literature) have discussed the technicality of text with different mode of writing, and the conclusion drawn lead us to the process of text unification towards all the data collected from the web and the other sources like newspapers, magazine and manuscripts of Maithili language. However, the alternations are found at various grapheme and pronunciation levels. Some problematic issues from the analysis have cropped up like how these varieties can be brought under unified approach and reconstruct the data for processing through technology. There are several examples which can justify not only the close class lexicons are added with huge numbers of scripting style but the open class as well. This reconstruction elaborates the issue hampering the way in unifying the script of Maithili Language. It’s messed up due to a variety of forms of Maithili, given to the socio-political reason especially in northern Bihar (Jha, Govind. et al.; 1986).