Linguistics Department Colloquia Series co-sponsored by the Center for Language Sciences

Daniel Siddiqi

Carlton University

513 Lattimore Hall, reception to follow

On Root Suppletion

Daniel Siddiqi (Carleton University)
Presenting co-authored work with Brandon J. Fry (University of Ottawa)

This talk aims to discuss the role of “root suppletion” in contemporary morphological theory.  The debate around root suppletion is largely about whether it even exists.  This question has proven to be resistant to being resolved as a simple empirical question.  Rather, it has been approached as a taxonomical question:  what are the criteria we have for considering a particular stem alternation to be suppletive?  Indeed, this is a central debate in the 2014 issue of Theoretical Linguistics--best exemplified in an exchange between Harley (2014a,b) and Borer (2014).  We summarize here Harley (2014a)’s argument from root suppletion, Borer (2014)’s proposal of criteria for assessing the existence of root suppletion. and Harley’s (2014b) response.

In this talk, we contribute to this debate by arguing that the search for root suppletion is better defined as “the search for counter-evidence to the L-node Hypothesis and the Early Root Insertion Hypothesis.” The L-node Hypothesis a prominent hypothesis in Distributed Morphology that claims that Roots are not individuated in the syntax (see for example Harley & Noyer 2000; Marantz 1996, Harley 1995, Harley & Noyer 1999, 2000).  The Early Root Insertion Hypothesis is a competing hypothesis in DM and similar models (such as Exo-Skeletal Syntax) that claims that Roots are individuated phonologically (see for example Embick 2000 et seq, Embick & Noyer 2007, Borer 2013).  Counter-evidence to these to hypotheses are taken to support the Root Competition Hypothesis, where Roots are individuated through some other means and participate in competition for Vocabulary Insertion (see for example Siddiqi 2009, Chung 2009, Harley 2014, Haugen & Siddiqi 2013, 2016).  Once we articulate the search in these terms, Borer’s (2014) criteria can be re-articulated such that they clearly show that the Uto-Aztecan data presented in Harley (2014) Haugen & Siddiqi’s (2013) clearly do not meet that criteria for falsifying these hypotheses.

Given the above, we present two sets of data from Ojibwe that appear to be root suppletion.  We argue that one set of data fails to meet these criteria, thus illustrating the importance of the criteria.   Meanwhile another set clearly and unambiguously meets the criteria, providing the sought after counter-evidence.  In this context, the question of sample size becomes relevant (as Borer 2014 argues).  How many examples of this type of root suppletion constitutes enough to falsify?  We argue that this is a metatheoretical question that needs to be evaluated against concerns of elegance, economy, and parsimony.  Finally, we argue that the types of root suppletion that are presented in Harley (2014a,b) serve an important function despite failing to falsify competing hypotheses.  We argue that are evidence confirming crucial predictions of the Root Competition Hypothesis.