Is English 'n't' a clitic or an affix?

Zwicky and Pullum argued that 'n't', as in 'isn't', behaves like an inflectional affix rather than a clitic, because it attaches selectively to auxiliaries and shows idiosyncratic forms such as 'won't'. The contracted auxiliaries like ''s' are clearer clitics.

What is a second-position clitic?

A second-position, or Wackernagel, clitic must appear after the first word or constituent of its clause, regardless of that element's syntactic role. Such clitics are common in languages like Serbo-Croatian and were central to debates about clitic placement.

Clitics and Affixes

Clitics are elements that lean phonologically on a host yet behave syntactically like words; distinguishing them from affixes on one side and independent words on the other is a central interface problem.

Definition

A clitic is a morpheme that is phonologically dependent on an adjacent host word but exhibits syntactic properties of an independent word, occupying a status intermediate between a free word and a bound affix.

Scope

This topic covers clitics and their relation to affixes and words: the diagnostics that separate clitics from affixes, the distinction between simple and special clitics, second-position (Wackernagel) placement, and the competing analyses of clitics as phrasal affixes or syntactic atoms. It does not cover word formation in general or the broader lexicalism debate, which are treated in sibling topics.

Core questions

How can clitics be distinguished from affixes and from independent words?
What are simple clitics, special clitics, and second-position clitics?
Why do many clitics appear in a fixed position such as second in the clause?
Are clitics best analysed as syntactic words or as phrasal affixes?

Key concepts

clitic versus affix
simple versus special clitics
second-position (Wackernagel) clitics
host selectivity
phrasal affix
clitic cluster

Key theories

Diagnostics for clitichood: Zwicky and Pullum's criteria distinguishing clitics from affixes, including selectivity of host, arbitrary gaps, morphophonological idiosyncrasy, and ability to attach across word classes.
Clitics as phrasal affixes: Anderson's analysis of special clitics as phrasal affixes whose placement is governed by morphological rules sensitive to phrasal edges, rather than by ordinary syntactic positioning of words.

History

The clitic-affix distinction was sharpened by Zwicky and Pullum (1983), who proposed influential diagnostics using the English negative 'n't'. The placement of special clitics, especially second-position effects first noted by Wackernagel, drove theoretical work on the interface. Anderson (2005) developed an account of clitics as phrasal affixes, and Spencer and Luís (2012) surveyed the typology and competing analyses.

Debates

Syntactic atoms versus phrasal affixes: Whether special clitics are positioned by the syntax like words or attached by morphological rules to the edges of phrases, a question that probes the morphology-syntax boundary.

Key figures

Arnold Zwicky
Geoffrey Pullum
Stephen R. Anderson
Andrew Spencer

Seminal works

zwickypullum1983
anderson2005
spencerluis2012

Frequently asked questions

Is English 'n't' a clitic or an affix?: Zwicky and Pullum argued that 'n't', as in 'isn't', behaves like an inflectional affix rather than a clitic, because it attaches selectively to auxiliaries and shows idiosyncratic forms such as 'won't'. The contracted auxiliaries like ''s' are clearer clitics.
What is a second-position clitic?: A second-position, or Wackernagel, clitic must appear after the first word or constituent of its clause, regardless of that element's syntactic role. Such clitics are common in languages like Serbo-Croatian and were central to debates about clitic placement.