Why can I form 'googleable' but not 'arrivement'?

The suffix '-able' is highly productive and attaches freely to verbs, whereas the pattern behind older nouns in '-ment' is largely unproductive in modern English, so speakers coin '-able' words but not new '-ment' words.

Blocking is the prevention of a new derived word by an already-existing word with the same meaning. For example, the existence of 'stole' blocks a regular form like 'stealed', and the existence of 'thief' can block 'stealer' in the relevant sense.

Productivity in Morphology

Morphological productivity is the extent to which a word-formation process can be used to coin new words, ranging from fully productive patterns that apply freely to unproductive ones surviving only in fixed forms.

Definition

Productivity is the property of a morphological process whereby it is available for the formation of new words; a process is productive to the degree that speakers use it to create and accept neologisms.

Scope

This topic covers the notion of productivity in word formation: how it is defined, why some affixes and patterns readily yield novel words while others do not, the role of restrictions and rival patterns, and the corpus-based and experimental methods used to measure it. It does not cover the units of morphology or the inflection-derivation contrast as such, though productivity is most often discussed for derivational processes.

Core questions

What makes one word-formation process productive and another unproductive?
How can productivity be measured quantitatively from corpus data?
What blocks or restricts the application of an otherwise productive pattern?
Is productivity a graded property or an all-or-nothing one?

Key concepts

productive versus unproductive process
blocking
hapax legomena
type and token frequency
rival affixes
neologism

Key theories

Word-formation rules and blocking: Aronoff's account in which words are formed by rules that may be productive to differing degrees, with the existence of an established synonym (blocking) preventing the coining of a rival form.
Corpus-based productivity measurement: The quantitative approach, developed by Baayen and colleagues, that estimates productivity from the proportion of hapax legomena (words occurring once) bearing an affix in a large corpus.

History

Productivity was foregrounded by Aronoff (1976), who introduced word-formation rules and the notion of blocking within generative morphology. Subsequent debate concerned whether productivity is a categorical or scalar notion. Baayen and Lieber (1991) pioneered corpus-based measures using hapax counts, and Bauer (2001) provided a book-length synthesis distinguishing the several senses in which a process can be called productive.

Debates

Is productivity gradient or categorical?: Whether productivity is best modelled as a continuous scale of probability that a pattern extends to new forms, or as a discrete distinction between rules that apply and lists that do not.

Key figures

Mark Aronoff
Laurie Bauer
Ingo Plag
Harald Baayen
Rochelle Lieber

Seminal works

aronoff1976
baayenlieber1991
bauer2001

Frequently asked questions

Why can I form 'googleable' but not 'arrivement'?: The suffix '-able' is highly productive and attaches freely to verbs, whereas the pattern behind older nouns in '-ment' is largely unproductive in modern English, so speakers coin '-able' words but not new '-ment' words.
What is blocking?: Blocking is the prevention of a new derived word by an already-existing word with the same meaning. For example, the existence of 'stole' blocks a regular form like 'stealed', and the existence of 'thief' can block 'stealer' in the relevant sense.