This book addresses a number of issues that originate from the various word orders found in the Bangla DP, assuming as an overarching framework the account of Cinque (2005) with respect to the universally possible word orders in the DP.

The first strand of argument deals with the leftward movement of the adjective to a post-demonstrative and pre-numeral position, leaving the NP in either its merge position in the DP, or in a position above QP to which the Adjective-NP sequence may move. When the NP is left in its merge position, the moved adjective has to be focused, as a non-focused moved adjective leads to ungrammaticality. This leads to the proposal for the existence of a second Focus phrase, Focus2, in the post demonstrative position in the Bangla DP, which is in addition to an earlier proposal for a pre-Demonstrative Focus1 position in the Bangla DP (Syed 2012). The proposal for a Focus2 position is defended in the context of a comparable post-demonstrative focus position in other languages. Focus2 is also shown to be less restrictive than the Focus1 position in the kinds of adjectives that it can host.

The second thread of argument in this book centres on the various interpretations of definiteness derived from the merge or the moved position of the NP and the presence or the absence of the demonstrative in the Bangla DP. As noted in the Bangla DP literature, in a DEM-less DP the NP in its merge position expresses indefiniteness, and the NP raised above the numeral-classifier expresses definiteness. I suggest that the raising of the NP is an instance of topicalisation which in turn realizes its definite reading. I propose the existence of a topic position below the DET in the Bangla DP where the definite reading is achieved. This Topic position is below the Focus2 position. Given that Syed (2012) postulates a pre-demonstrative Topic position above the Focus1 position, my analysis and proposals point to the existence of two left peripheries in the Bangla DP, above and below the Demonstrative respectively, that I suggest parallel the clausal left peripheries of Rizzi (1997) and Jayaseelan (2001).

My argument that Topicalization of the NP is responsible for the realization of definiteness in the absence of a Demonstrative in the Bangla DP draws on a proposal by Schwarz (2009, 2013) that definiteness can be of two types: unique definiteness and anaphoric definiteness. Bangla seems to fall into this paradigm of definiteness. I suggest that the raised NP in the absence of an overt demonstrative is ambiguous between unique definiteness and anaphoric definiteness. I further show that unique definiteness and anaphoric definiteness are separately morphologically represented (respectively) by two demonstratives in the Bangla DP: the proximal demonstrative ei ‘this’ and the anaphoric demonstrative sei ‘that.’ A third, distal demonstrative oi ‘that’ can indicate either unique or anaphoric definiteness. Rather than consider these readings as lexical properties of these demonstratives, I suggest that the deictic demonstrative and the anaphoric demonstrative are merged in two separate syntactic positions. As a consequence of the availability of two syntactic positions, the raised NPs (in the absence of a DEM) show the two readings of unique and anaphoric definiteness, as a reflex of their moving to and being licensed in these two distinct definiteness positions, when they superficially appear to move to a mere pre-Numeral surface position.

The third strand of argument in this book concerns the syntactic behaviour of the nominal and the modifying adjectives in ellipsis in the Bangla nominal domain. I show that Bangla conforms to Cinque’s (2012) generalisations regarding ellipsis that elements are elided only in their moved positions, and that DP ellipsis crucially involves movement and ellipsis of the N(P); with adjective or modifier ellipsis proceeding hierarchically from the kinds of movement permitted in the DP. In Bangla the plural classifier-gulo requires the NP to cliticize to it to its left. Thus the NP cannot stay in its merge position in the context of -gulo. Once the NP has moved the adjective cannot be left stranded (this is shown to follow from the parameters of movement in the Bangla DP, following Cinque (2005)). The NP moves pied-piping the adjective above the classifier -gulo. Both of them can get elided in their moved position. However, the adjective can surface even if the N(P) is elided, provided the adjective is focused. To escape elision, the adjective must move to the Focus2 position; it cannot move to the Focus1 position when the N is elided. In this connection some puzzling differences are noted between Adjective focalization with and without NP ellipsis; the latter, but not the former, allows violation of the Adjective Ordering Restriction. I also note instances of ellipsis apparently in the absence of NP movement, and try to subsume these instances under the more general account.

The final strand of argument of this book deals with wh-movement in the Bangla DP. I show that the wh word kon ‘which’ in Bangla merges in the post-numeral position and obligatorily moves to the pre-numeral position in the DP. I argue that the licensing position of the wh kon is not in the DP-initial position. This argument parallels Bhattacharya and Simpson’s (2000, 2003) argument (for wh-movement in the Bangla C-system) and Jayaseelan’s (2001, 2004) argument (for wh movement in the Malayalam I-system), that the moved wh-phrase does not occur in clause initial position.