In this paper, an attempt is made to put to text analytic and stylistic test Norman Blake's hypothesis (Blake, 1990) that the system of clause in literature-text has evolved from the eighteenth century all the way into the late twentieth in terms of noun groups becoming less complex and elaborate. A corpus of novels is selected such that every time span of fifty years can be represented by at least two novels. Then, in order to study the noun groups of respective evolutionary periods in question, it is the noun group in subject position that is subjected to analysis in terms of complexity, the key term of this stylistic study, which is taken to be analyzable through two yardsticks: a) the straightforward lexical density, ie the number of words inside the subject, and b) how many instances of ‘embedding’there are in that nominal functioning as subject. The study could not find support for Blake’s initial hypothesis as to the downward evolutionary trend in the complexity of noun groups, yet it brings out other implications and functions attributed to noun groups in subjecthood, over these respective historical periods, as used by novelists thereof.