Talk A - Drift, Markedness and the Final-Over-Final Constraint

Add to your list(s) Download to your calendar using vCal

• Alastair Appleton, Department of Linguistics
• Monday 22 February 2010, 17:00-18:00
• GR-05, English Faculty Building.

If you have a question about this talk, please contact Louise Radok.

This paper takes as its starting point the well-known observation that the word order change OV>VO is more common and seemingly more ‘natural’ than the change VO> OV (Newmeyer 2000). Indeed, Li (1977) claims that all non-contact induced changes are OV>VO. While Newmeyer reviews various functionalist explanations for this drift, I put forward a theoretical account for why internally-driven change might be expected to occur only in this direction.

I first adopt Kayne’s (1994) Linear Correspondence Axiom [LCA], meaning that that rigidly head-final (OV) languages must be derived by leftwards movement, or “roll-up”. In the simplest case, the complement (e.g. a DP object) raises past the head (e.g. V). This movement can be driven by an EPP -feature associated with the c-selectional feature of a head – i.e. V [ __ D, EPP ]. Following Roberts & Roussou (2003) and Roberts (2007) respectively, I also adopt the notions that (i) movement is a marked option and systems with more EPP -features are more marked than those with fewer such features; and (ii) the unmarked option is for all heads to have EPP -features iff V also carries an EPP -feature, with no EPP -feature being the unmarked option otherwise. We thus expect rigidly head-initial languages to be maximally unmarked systems, followed by rigidly head-final, then languages with successively more instances of roll-up movement.

Finally, I assume that Biberauer, Holmberg & Roberts’s (2009) Final-Over-Final Constraint [FOFC] holds. Informally, FOFC states that a head-final phrase may be dominated by a head-initial or head-final phrase, while a head-initial phrase can only be dominated by another head-initial phrase. If head-final structures are derived by movement, this observation can be captured if we require lower heads’ c-selectional features to carry EPP -features in order for higher heads to do so. This has consequences for change: if a language switches from rigidly head-final to head-initial, it must lose EPP -features from the top downwards. Feature loss starting in some intermediate position would generate a head-initial phrase dominated by a head-final one, violating FOFC . Once the topmost EPP -feature is lost though (presumably by grammaticalization of complementisers – cf. Roberts & Roussou 2003, ch. 3), we are left with a highly marked system. We might expect a pressure to reduce the markedness of the system, and so the remaining EPP -features are lost resulting in VO order. The three notions above thus conspire to provide an account of endogenous OV>VO drift, with no corresponding notion for exogenous change.

This talk is part of the Linguistics PhD seminars series.

This talk is included in these lists:

• All Talks (aka the CURE list)
• GR-05, English Faculty Building
• Linguistics PhD seminars

Note that ex-directory lists are not shown.