ABSTRACT
This chapter details the reach and complexity of American television. It also details a case study which focuses on the possibility of discriminating among news registers on television based on their verbal language alone. Linguists who conduct variation studies using the multidimensional framework face the challenge of building a corpus that is balanced and representative of the domain and language they want to investigate. A multidimensional analysis corpus should have at least five texts per variable being investigated. Thus, the first difficulty in corpus collection is identifying and selecting the registers the linguist is going to use from a multitude of television genres named, inconsistently, through the literature, listings, and the Television Academy, to build a large enough corpus for the study. Television genres have always been hard to define and classify. This difficulty stems from the fact that cultural products tend to have a “multiple generic participation” grounded on the varying perspectives of their different participants.
