South Asian Languages (from families as diverse as Indo-Aryan, Dravidian, and Tibeto-Burman) show a preponderance of verbal constructions usually labelled Complex Predicates (both the compound verb and the non compound verb constructions are labelled with the same tag) (see Vale 1948, Hook 1974, Dasgupta 1977, Mohanan 1994, Butt 1995, Gopabandhu 1992 among others). This study focuses on complex predicates (CPs) in Hindi-Urdu (HU) which are said to have Verb1-Verb2 (V1-V2), Noun-Verb (N-V), Adjective -Verb (A-V) structure. Focusing largely on ‘V1-V2’ and ‘N-V’ types of constructions as taken up in works such as Hook (1974, 1991), Butt (1995, 2010) and Mohanan (1994, 2007), I show how the ‘Verb’ in both types of constructions (V2 in ‘V1-V2’ and V in ‘N-V’), otherwise labelled a ‘Light Verb’ by many scholars, behaves differently with respect to phenomena such as reduplication, echo-reduplication and indirect causativization and argue that this allows us to reach a better understanding of the domain of their generation and the role of lexicon, morphology, and syntax. The crucial insights for this study come from the claims made by Butt (1995, 2010) regarding Aspectual Light Verbs (ACV) in Lahori Urdu which this study seeks to argue against in the light of evidence from the HU spoken in northern India. The study argues for the difference in the behaviour of the ACVs (Butt 1995) or the Vectors (Hook 1974) in the V1-V2 types of constructions and the light verbs found in the N-V type. This study is empirically grounded and argues the case based on data from several novels and short story collections of various writers of Hindi-Urdu spanning the whole of twentieth century, draws on speaker insights collected through questionnaires and google searches done during the period of this study. First-Phase Syntax (Ramchand 2008), Lexical Functional Grammar approach (Butt 1995), and Substantivist Linguistics approach (Dasgupta et al 2000) are some of the theoretical frameworks that were initially studied for insights. The study ends by proposing that the label ‘light verb’ be restricted to the ‘Verb’ in the N-V (and the A-V) constructions and that the label ‘Vector’ be strictly used to refer to the V2 in the V1-V2 constructions as they are definitely grammaticalized components as suggested by Hook in several of his works and as shown by the evidence presented in this study.