Word-formation: some general issues (Origins chapter 11)



The Oxford English Dictionary is online

§       On campus, use http://www.oed.com/

§       From home, go through the library homepage (e-resources: Oxford English Dictionary Online) or trick your server into thinking it’s at U of T


How does a language extend its vocabulary?

·       borrowing

·       most important in 16th and 17th centuries (Nevalainen 351)

·       brought in new affixes as well as new words

·       Most commonly, its own resources: wordformation (and deformation)

·       deformation: clipping, blending

·       formation: compounding, affixing, and functional shift (e.g. using noun as verb)

·       now (and except during that period in EModE) the most productive

·       particularly characteristic of modern English: compounds from classical elements that were never combined in classical times

·       Tyrannosaurus rex: Greek “tyrant” and “lizard”; Latin “king”


What is the difference between compounding and affixing?

·       compounds are formed from two free-standing words (“independent lexical items”): fleabite, bookseller (C16th)

·       affixing involves adding an affix to a free-standing word: non-toxic

·       prefixes tend to change meaning (toxic!)

·       suffixes tend to change word class (toxicity)

·       borderline: neo-classical formations like astronaut

·       can’t stand alone: electro-, -phile

·       but can form entire words: hydrophobe




·       combining two freestanding lexical units to form a new one: studmuffin, brewpub

·       common in OE

o      including loan-translations (or calques): two elements of Latin patri-arch recognized and rendered with two OE elements, heah-faeder

§       workbook 12.4 has some more modern ones

o      creative: hláf-weard ‘loaf-ward’, wif-mann ‘female human’

o      lord and woman

o      subsequent sound changes sometimes impede analyzability

o      Algeo’s “amalgamated compounds”

o      workbook 11.13


In EmodE, most compounds are original

o      loan translations include fairy tale 1750 on conte de fée


Most common compounds: N+N

o      different kinds of relationships between the Ns

o      e.g. ‘copulative’: maidservant, girlfriend; washerwoman (servant, friend, washer who is a …)

o      e.g. ‘rectional’

§       e.g. with a ‘object-subject’ relation: boatman (man has a boat)

§       e.g. N2 is for the N1: ashtray, doghouse

o      workbook 11.12 has more

o      lack of inflections can obscure the relationship: teen killer (anecdote: coin laundry)


Stress patterns useful for establishing

§       meaning: TEEN killer, teen killer

§       lexicalization: white house vs White House


If NN phrases do become lexicalized, they can get clipped

§       (computer) programmer

§       cell(ular) (phone)


Foreign patterns

§       governor general, court martial

§       not productive




Prefixes: usually change meaning not class

§       un/happy

o      interesting form: inflammable (does in- negate or intensify?)

Suffixes: usually change class

·       happi/ness

§       not always: sonneteer, rhymester


Affixing in EmodE:

Native affixes are most high-frequency

§       blurry area between vocabulary and grammar: verbal nouns in –ing

§       most common noun affixes

o      -er derivations (forms nouns from verbs): sinister whisperers

o      -ness: makes abstract nouns from adjectives (goodness)

§       “prefers native bases but isn’t limited to them”: disingenuousness


But some foreign ones became very productive in this period

§       –ness and –ity are the main suffixes that derive abstract nouns from adjectives

o      –ity mostly sticks to Latinate bases (oddity an oddity)

o      but there is some overlap with –ness, and it’s very productive

§       helps to look at bigger context


What is productivity?

·       “the capacity of a word element ... to produce new words”

·       the amazing Brian-iz-er

·       unputdownable

·       in contrast to the unproductive transparency or analyzability of –th

·       recognize it as having formed abstract nouns like truth, warmth, growth, length, breadth, width

·       sloth isn’t analyzable/transparent because of subsequent sound changes

·       but doesn’t make new words: the few new formations are jocular: coolth, thickth


What might block new formations?

·       phonological characteristics of the base form

·       we don’t usually make adverbs out of adjectives in –ly (friendlily tends to be avoided)

·       existing synonyms “preemption by synonymy”) (McMahon 195): domestic blocks housely or housey; vernal blocks springly (Gorlach 172)

·       existing verbs in ify or ize might block new ones in ate: no edificate, deificate, pulverizate

·       weren’t necessarily such norms in EModE before dictionaries: verbs attested glad, gladden, englad, engladden and beglad (Gorlach 172)


Elimination of redundancy after EModE

-principle: languages tend not to have forms that are identical in form, meaning, register, region, etc.: something has to differ

-verbs in –ate often competed with their infinitive form

-variants later eliminated or redistributed

-expire, expirate: disappears

-prove, probate: semantic differentiation

-orient, orientate: regional differentiation


How do foreign affixes become productive?

·       first, borrowed words appear that are complex, i.e. that contain affixes

·       e.g. charioteer and engineer

·       then the affixes become analyzable: “person concerned with or engaged in, often in a military context”

·       then the affixes might start to be productive

·       cannoneer, volunteer: often military

·       pulpiteer, pamphleteer: extended (with ironic hyperbole) to non-physical combat

·       waistcoateer ‘prostitute’, profiteer: or just derogatory without denoting combat

·       mountaineer, auctioneer: though not always derogatory

·       mostly attached to Romance bases


Some affixes have the ‘same’ form but different origins

§       see workbook 11.7: amoral and aside



-English forms new words by adding affixes: sleep, sleepy, sleeper

-occasionally the reverse happens: a simple form is wrongly analyzed as a complex form, and a supposed ‘suffix’ is removed

-see workbook 11.16 for examples

-because –s is the productive plural inflection, it’s sometimes removed from singular loanwords

§       French cerise -> cherry

§       Greek heros -> hero

§       Dutch schaats -> skate


          -agent noun suffix –er added to verbs to make nouns

-so Latin editor -> edit, sculptor -> sculpt (though see OED)

                   -also cobble, tipple from cobbler and tippler


-adjective ending –y very productive: pattern of sleep and sleepy spawned new nouns and verbs

-new nouns greed (1609) and haze (1706)

-later: nouns sleaze and funk from sleazy and funky

-verb laze formed from lazy (16th century)


-enough nouns in –ation corresponded to verbs in –ate that new verbs were formed from existing nouns in -ation

-legislate (c18th) from legislation, negate (17th century) from negation

-orate and donate: borrowings from Latin? or back-formations from oration and donation?


-some modern verbal nouns in –ing seem to have appeared before the corresponding verbs: multitasking, dumpster-diving


-same with other verb/noun pairs

-revise/revision model for television -> televise

          -liaison -> liase


-tend to be semantically derivative (televise ‘to put on television’, rather than television as conceived as an act of televising)



“Classical compounds”

o      Latin and Greek words were being adapted in the European vernaculars as scientific and technological terms, e.g. virus, vacuum, etc.

o      so were elements of Latin and Greek, often newly combined

o      hybrids combine Greek and Latin:

          anti- (Greek, ‘opposite’) toxin (Latin, ‘poison’)

tele (Greek, ‘far’) + vision (Lat, ‘seeing’)


o      note: spelling of Greek often depends on where we got it from

          -e.g. Greek /k/

                   -<k> if directly from Greek, e.g.kinetic

                   -<c> if from Latin or French, e.g. calligraphy

                   -<c> /s/ if sound change within French, e.g. cinema



Morphological status of classical compounds


o      elements like bio- and –ology like dis- and –ness in that they can’t stand alone

o      but they can combine with each other (*dis-ness)

biocide, ecophile, astrology



Etymological status of classical compounds

o      Would you say television is Greek, Latin, or English?

o      not necessarily coined in an English-speaking country

o      present in most languages, e.g. antitoxin is recognizable in English, French, Italian, Swedish, Russian (calqued in German: Gegen-gift ‘against  poison’)

o      ways around it: “International Scientific Vocabulary”, just cite elements (OED) …


Occasional reanalysis

-auto “self-“ -> “pertaining to cars” in Autoshare

          -but still has meaning of ‘self’ in, e.g. autoimmune disease

-tele “far” -> “pertaining to televisions” in telegenic



Functional shift (or conversion or zero-derivation)


-very common: using one part of speech as another (foreground ‘to put in the foreground’)

-because English now has so few grammatical endings

          -OE lufu, lufian -> PDE love


-many noun -> verb: input, foreground, trash, impact, snowboard, mountain bike

          -some the other way: double-click was a verb before it was a noun


-most share stress patterns: is this one way of telling whether a verb has come from a  noun or not?

-Marchand suggested that when there is a difference in stress, the noun has come from the verb

          -noun |address comes from the verb ad |dress


-but we also find difference in stress when the verb has come from the noun: escort (noun 1579, verb 1708), progress (noun 1432, verb 1590)


Some the result of several processes in sequence

-private ‘lowest rank in army’ became a noun from an adjective as a result of the clipping of private soldier



More “modern” processes:


Eponyms (derivation from common nouns)

-sandwich, stetson

-kind of metonymy (the thing || the inventor)

-usually commercial

          -can be short-lived, if the thing disappears: Hansom, Anderson


Root creation

o      hard to document earlier

o      lots of onomatopoeic words: clank, splash

o      words like fop, fib have no known antecedents


o      easier with commercial or scientific trail! Sound symbolism important

o      Blurb, Kodak, quark, nylon

§       -on or –lon now has a life of its own

§       Teflon: Tef sounds like ‘tough’, -lon something smooth and high-tech

·       Exploits association with neo-Greek –on (electron, proton, etc.)


Often not utterly made up

-Algeo: even when companies do it with random computer generation, humans choose from the results

          -e.g. CVCVC + suffix like –el, -ex, -on



-unique, novelty (both science and commerce)

-accessibility (AAAA locksmiths)

-memorable sound

-no unfortunate semantic suggestions, e.g. Nova a bad name for a car in Spanish




-different processes, but basically terminal loss in first item and initial loss in second item, where there are usually overlapping parts in the fusion

-some from earlier: twirl ‘twist’ + ‘whirl’? 1598, blotch from blot and botch?

-Carroll’s chortle (chuckle + snort), galumph (gallop + triumph) almost established

-many quite deliberate

-commercial: motor + hotel (botels in Florida), selectric, infomercial

-science, news (Watergate, Gatesgate, Spraypec) … shows short-lived nature of some


Clipping: app ‘application’, fax ‘facsimile’

-informal language: slow to appear in writing

          -but C16th examples: chap, gent, coz from chapman, gentleman, cousin

          -cute from acute 1731 (foreclipping)

          -miss from mistress 1666 (backclipping)

-sometimes part of adaptation process

          -spinet, special, spy from French espinet, especial, espy

          -but unstressed first syllable also lost from native words: a/lone, a/live in C16th

-can change stress: examination to exam

-can change meaning of what’s left: soap, cell, chauvinist, auto, bus

          -chauvinist ‘blind patriotism”

          -bus: just part of a Latin grammar ending in word omnibus (vehicle for all)

-often cuts off the end of a word or phrase

-but not always: bus, fridge, flu

-and not at the original morphological boundary

                   -strep throat (strepto-coccus)

                   -condo (con-dominium)

-can lose etymological spelling: fax from facsimile



-acro ‘tip’ + nym ‘name’

-term coined by analogy with ‘homonym’


-extreme kind of clipping

-relatively recent: need literacy for it

          -we know that people pronounced a.m, p.m., M.A.



          -efficient, authoritative

          -catchy: MADD

          -euphemistic (phonetic distortion):

general use: BO ‘body odour’

medicine: TB/AIDS, HIV

war: HRP (“human remains pouch”)

          -often lose the origins

                   -e.g. laser – I’ve given up trying to reconstruct it


Algeo distinguishes

-alphabetism (initialism): UN, CBC, etc.

-acronym: pronounced as single word NATO, UNESCO, RAM

          -can use first few letters in order to make them more pronounceable

-so trendy that there are ‘reverse acronyms’: start with the result you want, like MADD

                   -“a sort of political offshoot of normal acronymic coinage”



Appendix: case studies of some productive affixes (with thanks to ZP)


Suffix –eer an Anglicization of French –ier (itself from Latin –iarius > -ary)

·       forms nouns from other nouns

·       sense “person concerned with or engaged in”

o      except for a few words like gazetteer and muffineer “device for putting sugar on muffins”


First found in loanwords from French in the C16th

·       interchangeable with –ier: muletier/muleteer (and tons of other forms: -er, -our...)

·       though charioteer and engineer come in earlier (ME engineour) and get assimilated to –ier, -eer


In C17th, -eer and –ier become distinct: brigadier, grenadier, gondolier


And –eer starts to be productive within English

·       privateer “armed vessel (and its commander) owned and officered by private individuals holding a government commission and authorized for war service”

·       though mostly with Romance forms?


Many early examples have military connotations: cannoneer, volunteer


Extends to non-physical forms of combat: pulpiteer, pamphleteer

o      in derogatory contexts

o      ironic hyperbole?


And can be derogatory without denoting combat : waistcoateer ‘prostitute’, profiteer


But some words are neither derogatory nor agonistic: auctioneer, mountaineer


Consider contexts

·       mostly Romance bases (cf. –ster: rhymester, fraudster, etc.)


Tends to occur with –ing (and occasionally to be backformed from these forms)

·       engineering, pamphleteering

·       parliamenteering, revolutioneering (OED has no entry for revolutioneer)



Appendix: verbal -ate

·       initially “an anglicizing termination with Latin participles”: creare ‘to create’, creatus ‘created’

·       -ate first appeared in Middle English in participial forms adapted from the second participle of the Latin first conjugation (creare “to create”, creatus “was created”; there were also related nouns like creatio, creationem, anglicized as –ation)

·       ex: “Man was create”

·       ex: “The opponents were separate”

·       cf. adjectives like inveterate, immediate

·       One theory for early lack of ed: reinforced by other paradigms with infinitives identical in form to participles or adjectives:

·       Romance words like clear: CLEAR the table so that it is CLEAR

·       Native verbs like cut: CUT the fabric so that it is CUT

·       Native verbs like dry: DRY the dishes so that they are DRY.

·       “after about 1400, appeared with other verb forms”, e.g. infinitives with causative meaning

·       ex: “He saw St Peter consecrate him king” (1387, supposed to be earliest instance of a full verb in ate)

·       the past participle consecrate is first attested around the same time

·       ex: “God will create man”

·       ex: “They will separate the opponents”

·       once there were enough pairs

·       you didn’t need to have the participle come in first

·       the present/infinitive form of the loanword can come first

·       the suffix ­­–ate  become morphologically transparent: base and a separate meaningful affix

·       you can make up new words within the language

·       usually then combines with borrowed bases

·       “from the sixteenth century onwards, used to form verbs from

·       Latin nominal stems: Latin pagina -> paginate; facility -> facilitate

·       and Romance bases: adjective vaccine (“vaccine disease, cow pox”) -> vaccinate

·       or from French vacciner (French –er || -ate)?

·       it’s usually impossible to tell whether new forms are in fact back-derivations from nouns in –ation, which are often older than the verb

·       e.g. is American orate back-formed from oration or directly taken from Latin orare

·       appears intermittently in OED since 1600 (earlier occasions more likely from Latin)

·       “I denounce the allegations and I denounce the allegators”

·       Why is this funny? (Back-forming allegate from allegation and then deriving allegator from it)

·       now many verbs that never existed in Latin

·       some not scientific:

·       hyphenate, orchestrate

·       very productive in scientific English

·       chlorinate, dehydrate

·       (likes to combine with de-): defibrilate

·       or in pseudo-scientific or slangy English

·       garburator (would have to back-form the verb)

·       absquatulate, discombobulate