Skip to navigation – Site map

HomeAnglophonia30The suffix -ee: history, producti...

The suffix -ee: history, productivity, frequency and violation of stress rules

Ives Trevian

Abstracts

According to Bauer (1983), Barker (1998), Plag (2003) and Mühleisen (2010), -ee has become a very productive suffix, not confined any more to its original role in legal language as a patient suffix contrasted with the agent suffix -or (e.g. bailor / bailee). Many new coinages are supposed to have appeared in the 20th century, so that 163 words in -ee are now recorded in the online dictionaries of the OneLook search engine (e.g. American Heritage Dictionary, Cambridge Advanced Learner’s Dictionary, Collins English Dictionary, Merriam Webster’s Dictionary, Oxford Dictionaries, etc.). Some of the words formed with this suffix are now agent instead of patient nouns (e.g. escapee, returnee), occasionally rivalling with other agent suffixes (e.g. signee = signer). Moreover, besides the traditional deverbal formations (e.g. abductee), denominal and even deadjectival coinages are now sanctioned (e.g. asylee, presentee). The purpose of this paper is (1) to show how this historical process came into being, namely when the change occurred from usage in legal language to other uses; (2) to examine this suffix in terms of its morphological productivity, notably to distinguish well-established nouns from mere nonce words, chiefly used as strictly contrastive (e.g. cutter / cuttee, jester / jestee); (3) to ascertain the frequency of -ee suffixed nouns; to do so we have confronted our corpus of -ee nouns with the word frequency search engine Google Books Ngram Viewer and the Collins Dictionary’s Recorded Usage Charts, the former noting the frequency of words since the 16th century and the latter containing dated sentence examples; (4) to determine how many words in -ee violate the rule proscribing two consecutive stresses (e.g. aˌsyˈlee, aˌwarˈdee), this violation being, in principle, allowed only in compound words (e.g. ˈbootˌlegger, ˈteenˌager) or semantically transparent formations with a prefix (e.g. ˌreˈmake, ˌunˈthaw, verbs); an attempt to account for these violations is proposed in this paper, which has inevitably led us to ponder over the impact of this suffix on metrical rules.

Top of page

Full text

1. Introduction

  • 1 For non-sapient nouns like bootee = short woolen socks that babies wear instead of shoes” or goa (...)

1The suffix -ee has had a long career in English since the Norman Conquest and up to the present day. An adaptation of Anglo-Norman, it has to be defined morpho-semantically as a morpheme denoting a human being1 and attaching to an identifiable base, the resulting meaning of the suffixed form being transparently deducible from those of both constituants (detainee < detain, verb + -ee).

2 Syntactically and semantically, -ee was in Middle English attached to transitive verbs to form patient nouns denoting the recipient of an action (grantee < grant + -ee). Quite early, however, some of the words formed with this suffix became agent instead of patient nouns (cf. section 2, first paragraph) deriving this time either from transitive or intransitive verbs (e.g. returnee < return, intransitive sense) and occasionally rivalling with other agent suffixes (signee = signer < sign). Moreover, besides the traditional deverbal formations (e.g. abductee), denominal and even deadjectival coinages are now sanctioned (e.g. asylee < asyl(um) + -ee, presentee < present, adjective). These points will be further developed in sections 2 and 3.

  • 2 In his well-received book, Dixon (2014: 51) curiously states that the -ee suffix is attested in a (...)
  • 3 But much less so for other languages, notably Chinese. “Google n-grams and pre-modern Chinese” (n (...)
  • 4 Even though they can be interesting for linguistic analysis, notably for the identification of no (...)

3 The suffix -ee is supposed to be highly productive according to Bauer (1983), Barker (1998) Plag (2003) and Mühleisen (2010)2, who have identified many coinages dating from the 20th century. The question will be asked whether these authors, who chiefly based their dates of first attestation on those provided by dictionaries such as the Oxford English Dictionary (henceforth OED), have been right in this respect. It is also important to distinguish well-established nouns from mere nonce words, chiefly used as strictly contrastive (e.g. cutter / cuttee, jester / jestee). To ascertain the years when suffixed nouns in -ee were coined the word frequency search engine Google’s Ngram Viewer (henceforth NgV) is an excellent tool for English lexical research3 as it notes the frequency of words from a large corpus of texts (more than 20 million books) digitised as part of the Google Books project. Ngram is a measure of the likelihood that a given word will appear in a two-, three-, four-word, etc., string of words (Kies 2010: 2). It goes from the year 1500 through 2019 and thus gives the earliest recorded date of appearance for each word it contains. It also provides peak years of usage and is totally free and faster than electronic corpora such as the British National Corpus (BNC), in which only language use from the late 20th century is represented, or the Corpus of Contemporary American English (COCA), which covers the years 1990 through 2017. True enough, the latter two corpora can be used to find hapaxes (Bauer et al: 2013) but the approach in this paper is geared towards a corpus based on dictionaries, which has been extracted from the OneLook search engine. OneLook gives free access to well-established general-purpose dictionaries of the English language (e.g. American Heritage Dictionary, Cambridge Advanced Learner’s Dictionary, Collins English Dictionary (henceforth Collins D.), Dictionary.com, Merriam Webster’s Dictionary (henceforth Merriam Webster’s D.), Oxford Dictionaries, etc.), together with specialised dictionaries.4 163 words in -ee are now recorded in the dictionaries accessible from OneLook, including some of the 20th-century coinages collected by Bauer (1983), Barker (1998), Plag (2003) and Mühleisen (2010). These words will be studied and confronted with NgV in sections 3 and 4.

4 In terms of stress assignment things seem pretty straightforward as the transparent -ee suffix supposedly always takes primary stress. However, the assignment of secondary stress is subject to much variation, being apt to violate the rules of stress distribution in English. An attempt to account for these violations is proposed in this paper (cf. section 5).

2. A short history of the -ee suffix

5The suffix -ee was used “towards the end of the Middle English period, in the fourteenth and fifteenth centuries” (Mülheisen 2010: 62) as a correlative to the suffix -or in a legal sense to denote the sapient but non-volitional recipient of an action initiated by the form in -or (bailor / bailee, grantor / grantee, etc.). The suffixes -or and -ee were freely attached to transitive verb bases to form nouns, those in -or referring to the agent and those in -ee to the passive party (or patient), in transactions or legal deeds. First confined to legal language, the use of the suffix was, in the 18th and 19th centuries (Mülheisen 2010: 74), frequently imitated in the formation of “humorous (chiefly) nonce-words, as cuttee, jestee, laughee, denoting the personal object of the verbs from which they are formed” (OED). In such examples, the terms in -ee are generally coupled with agent suffixed nouns in -er (e.g. the cheater and the cheatee). Besides, as early as the 17th century, the -ee suffix was given a new lease of life and began being used outside legal language and jocular expressions (e.g. attendee (1679), retiree (1623), dates taken from NgV), which are still some of the most commonly used words in -ee, cf. Table 2 below). As can be seen from bargee (1674 < barge, noun), the deverbal formation of nouns in -ee was no more an ironclad rule. Since then, new coinages have steadily been added into the lexicon. As seen above (section 1, third paragraph), although they do not (yet?) appear in dictionaries, many more cases have been reported by Bauer (1983), Barker (1998), Plag (2003) or Mülheisen (2010), to the extent that Mülheisen has listed several hundred suffixed nouns in -ee, cf. section 3, i and ii).

  • 5 Plag (2003: 88), who does not adhere to Guierrian generalisations (e.g. -VʹVʹ(C)), simply describes (...)
  • 6 Contrary to Guierre (1984), Duchet (1991) and Abasq et al (2019), -aaC has been discarded as the (...)

6 In terms of stress, it is at this stage important to distinguish synchronically opaque items in -ee from words suffixed with the morpheme -ee, in which this suffix attaches to a transparent base. One of the rules on end-stressed words or -VʹVʹ(C) (Guierre 1984: 59; Duchet 1991: 11, 13; Carr 2012: 175; Abasq et al. 2019: paragraph 10 and table 6)5 is that “two identical vowels, followed by zero or more consonants”, take primary stress. The rule works very well for -oo (bamboo, etc.) and -oon (platoon, etc.) with very few exceptions (all in -oo, e.g. ˈigloo, ˈvoodoo). It also works for words in -eeC, whether the formation is transparent or not (careen, velveteen, career, pamphleteer, etc.).6 However, when analysing the data provided by OneLook general-purpose dictionaries, it appears that a clear majority of monomorphemic words in -ee do not abide by this rule, as shown in the examples below:

7Table 1. Stress patterns of monomorphemic words in -ee

[(-)10] or [(-)100]

'ackee (= “fruit”), 'apogee, 'banshee (+ variant [01]), 'bungee (= “stretchy rope”), 'chickadee (= “bird”), 'coffee, (ˌsub)com'mittee (different from ˌcommit'tee < com'mit, in the obsolete legal sense of “an individual to whom the care of a person or a person's estate is committed”), cor'roboree (= “native Australian language” < corroborate), 'fil(l)agree (= “orrnament”), 'filaree (= “plant” or -'ree), 'freebee, 'gidgee (= “acacia”), 'greegree, 'hackee (= “chipmunk” <≠ hack), 'jubilee (demotivated < jubilate), 'lychee, 'levee, ˌmaha'ranee, 'mallee, 'manatee (+ var. [201]), 'matinee (= [-neɪ] + [201] in US English), 'pharisee, 'pedigree, 'pewee, 'perigee, 'puttee (< Hindi <≠ put), 'ranee (spelling var. of “rani”), 'spondee, 'te(e)pee, 'toffee, 'toupee (or -'pee), 'towhee (= “sparrow”), 'trochee, 'weewee, 'whoopee, 'yankee (<≠ yank), etc.

[(-)01]

ˌchimpan'zee (US = chim'panzee), ˌdunga'ree (+ var. [102]),ˌguaran'tee (+ var. [102] in US), ˌjambo'ree, mar'quee, ˌrepar'tee (+ [-teɪ] in US English), etc.

  • 7 Other OneLook dictionaries may differ in the number of their definitions but they all set apart t (...)

8It is therefore necessary to correct the -VʹVʹ(C) rule as far as -ee is concerned to: suffix -ee bears primary stress whereas final -ee in monomorphemic words most often does not. Once monomorphemic words in -ee have been set aside, we are confronted with several senses of this suffix, as in Collins D.:7

-ee: suffix 1. indicating a person who is the recipient of an action (as opposed, esp. in legal terminology, to the agent, indicated by -or or -er): assignee, grantee, lessee; 2. indicating a person in a specified state or condition: absentee, employee; 3. indicating a diminutive form of something: bootee.

9The origins of the suffix differ. Thus, in 1. and 2. we have a loan from Old French -e, -ee past participial endings, from Latin -ātus, -āta, whereas the -ee in bootee, goatee, etc. is of uncertain origin, probably a spelling variant of -y, -ie, though stressed as if formed with -ee (Dictionary.com). Therefore, it cannot be linked semantically to 1. and 2. Although -ee as a diminutive also bears primary stress, I shall confine my analysis in this paper to the more productive suffixed forms denoting a person as in assignee, employee, etc.

3. Productivity and uses of the -ee suffix

  • 8 The dates of first written occurrences have been taken from NgV.

10The suffix -ee was very productive in the 20th century (Barker 1998), whatever the type of measurements considered, including that of hapax legomena as set forth by Baayen (1992, 1993), Baayan and Lieber (1991) and Baayen and Renouf (1996). Barker lists 74 pairs of the standard law terms of the grant(or/ee), vend(or/ee) family, then 96 nouns in -ee which supposedly joined the lexicon in the 20th century (Barker, 1998: 37), most being now recorded in general-purpose dictionaries (e.g. appraisee 1926, contactee 1910, counselee 1913, enshrinee 1945, retrainee 1934)8. Other scholars such as Bauer (1983), Plag (2003) and Mühleisen (2010) have given many more examples of recently coined nouns in -ee which, although not listed in OneLook dictionaries, are nonetheless attested in Internet sources (cf. paragraph below Table 2): blackmailee, congratulee, holdupee, huggee, slanderee, etc. (Bauer 1983: 248).

  • 9 Despite the etymology given in dictionaries, it cannot be excluded that biographee actually resulte (...)
  • 10 Although synchronically denominal refugee (< refuge) is apparently an anglicisation of refugié wh (...)

11 Fundamentally deverbal, the -ee suffix occasionally attaches to nouns as it did with bargee (cf. section 2, first paragraph): asylee (< asylum), biographee (“subject of a biography” < biography)9, mentee < mentor, refugee < refuge)10. Even deadjectival derivatives such as deadee (in the sense “portrait of a dead person painted from a photograph”, cited by Mühleisen (2010: 114), from Barnhart et al. (1990: 152), not available from OneLook), and presentee are licensed. Deadee is indeed recorded in NgV, its date of first attestation being 1803, and was still used in the year 2018, though with a very low frequency rate (0,0000000047%). As for presentee, it is deverbal (17th century) in the sense of “a person presented or nominated” or “a person to whom something is presented” but indeed deadjectival (19th century) in that of “a person who is present” (OED). It is also still used in present-day English (last occurrence 2018 in NgV).

12 Under the influence of American English (Huddleston and Pullum 2002: 1697), the -ee suffix has extended throughout the English-speaking world to other lexical fields, still with a passive sense (in association with transitive verbs) of someone affected by an action (abortee, auditionee, etc.) but also with the active sense (in association with transitive and intransitive verbs) of someone performing an action (e.g. signee, which rivals with signer in the sense of “a person signing a document”, escapee, which has displaced escaper in the sense of “someone escaping from”, returnee, derivable from ‘return’ in its intransitive sense, etc.). The -ee suffix can even rival with conversion as in conscriptee (1796) vs. conscript (noun, 1671). Both were still used in 2018 according to NgV. Once the infinitive of the verb (found by searching “to conscript”) has been deducted from all occurrences (0,0000277685% for conscript, noun or verb, minus 0,000000040653% for to conscript = 0,0000237032%), the noun conscript turns out to be much more frequent than conscriptee (0,0000000189%).

  • 11 As quite a few nouns from her list, like many in Mülheisen’s corpus, are absent from OneLook dict (...)

13 When combining with a verb in -ate, -ee is supposed to entail the elision of the verb termination (e.g. amputee < amputate, evacuee < evacuate, nominee < nominate, etc.). This truncation rule referred to by Bauer (1983: 247) is however, as shown in Anderson (1992: 280), disallowed in derivations from two-syllable verbs in -ate (e.g. mandate > mandatee or narrate > narratee) which, if truncated, would lead to unrecognisable stems (*mandee, *narree). The same reason explains the non-deletion of -ate in dedicatee < dedicate, educatee < educate, delegatee < delegate, etc., where removal of the verb termination would make the suffixed form hardly comprehensible (*dedicee, *educee, *delegee, Trevian 2015: 77). Indeed, before the graphic vowels e/i/y, graphic c and g imply respectively a realisation in [s] instead of [k] in most English words (e.g. decent, cinder) and [ʤ] instead of [g] in words of Romanic origin (e.g. divergent, regiment). It is, however, to be noted that in her extractions from web sources (cf. first paragraph below Table 4) Mühleisen finds many examples of what she calls “new forms (or “new words”) found on the Web” where -ee attaches to verb bases in -ate without entailing deletion of the verb termination (Mühleisen 2010: 2223): anticipatee (1828), animatee (1842), commemoratee (1856), compensatee (1880), congratulatee (1803), consecratee (1818), discriminatee (1825), dominatee (1847), separatee (1657), etc.11 Whereas, as shown by their dates of first written attestation in NgV, these nouns are definitely not “new words”, it cannot be denied that the truncation rule set out above has been quite chaotic in the history of English. Mülheisen is to the best of our knowledge the scholar who has written the most ambitious work on the -ee suffix. In her 2010 book she claims that:

i. in 2005, approximately 500 nouns suffixed with -ee had already been collected;
ii. her analysis of the Internet provides […] “evidence that the productivity of this word-formation pattern may be significantly higher than assumed, both in ad-hoc creations and nonce words, as well as in formations which seem to be already well established in usage but not documented in dictionaries” […]. (Mülheisen, 2010: 10).

  • 12 As said above, they resorted in most cases to OED for dates of first written attestations.

14From the studies by Barker, Bauer, Plag and Mülheisen mentioned above, it would seem quite safe to assume that -ee is indeed very productive in terms of morphological derivation. However, when looking at these authors’ lists of words which, according to them, appeared in the 20th century,12 NgV shows us that many of them were actually recorded as early as the 17th to 19th centuries, which nonetheless does not mean they are not still used, as shown in the examples below:

15Table 2. Dates of first written attestations (OED, Collins D. and NgV) and latest recorded use (Collins D. and NgV)

attendee
1961 in
OED vs. 1709 in Collins D., latest recorded use 2013: “They represented only 38 percent of attendees to the conference last year”, Times, Sunday Times; 1679 in NgV, latest recorded use 2019

detainee
1928 in
OED vs. 1722 in Collins D., latest recorded use 2016: “Thousands of long-serving political detainees had indeed been released from the grim Cuban gulag under his more reformist regime”, Times, Sunday Times; 1668 in NgV, latest recorded use 2019

franchisee
1956 in
OED vs. 1800 in Collins D., latest recorded use 2015: “The deal is open to would-be franchisees already running their own shop or those fresh to the business, as the firm will help find a suitable store”, The Sun; 1658 in NgV, latest recorded use 2019

inductee
1941 in
OED vs. 1765 in Collins D., latest recorded use 2007 (no example given); 1702 in NgV, latest recorded use 2019

internee
1915 in
OED vs. 1755 in Collins D., latest recorded use 2008 (no example given); 1603 in NgV, latest recorded use 2019

retiree
1935 in
OED vs. 1726 in Collins D., latest recorded use 2015: Why should retirees wanting the same income see their savings taxed at 55p in the pound?”, Times, Sunday Times (2015); 1623 in NgV, latest recorded use 2019

selectee
1940 in
OED vs.1804 in Collins D., latest recorded use 2007 (no example given); 1702 in NgV, latest recorded use 2017

testee
1932 in
OED vs.1818 in Collins D., latest recorded use 2007 (no example given); 1787 in NgV, latest recorded use 2019

16Some of them have of course changed meanings, which may explain why some authors consider trainee (formerly a “trained animal”) as a word dating from the 20th century in its present sense. Similarly, franchisee initially denoted “a person who had been granted a special privilege by a sovereign power” (OED) whereas, since the mid-1950s, the noun has taken on a more specific meaning (“a person to whom a commercial franchise is granted”). Quite a few nouns in -ee, culled by Bauer, Barker, Plag and Mülhensein, which are not recorded in OneLook dictionaries, are older than they thought (e.g. blackmailee (1892), huggee (1834), indicatee (1859), investigatee (1858), in NgV) whilst others, as said above, cannot be found in NgV (e.g. congratulee, relegatee, slanderee). This does not mean that -ee is not productive in terms of morphological derivation but there remains to be seen how many ad-hoc creations and nonce words, as Mülheisen puts it, are to be counted. Thus beseechee, discussed by the latter (Mülheisen 2010: 1) most certainly exists, even if it is not recorded in NgV, as it serves as a parallel paradigm to beseecheer in contrasted sentences. Of the 748 successful -ee words she tested “552 items (73.8%) occurred in at least one instance in ‘lexical solidarity’ with an -er word” (e.g. applauder / applaudee (not recorded in NgV), sneerer / sneeree (id.)), “whereas 196 items (26.2%) produced no such collocations” (Mülhensein 2010: 31). This entails that nearly three quarters of her confirmed corpus are ad-hoc or nonce words, in other terms “words coined for one single occasion only”.

  • 13 Accusee is listed as having existed since 1803 in NgV.

17 Another problem may arise with the -ee suffix, namely rejection by prescriptivists (or purists) of a formation. Thus accusee, which cannot yet be predicted to be on its way to replacing accused (as in ‘the accused’), is condemned by many, including Garner (2009: 16), who qualifies it as “a needless variant of accused”, although it is recorded as a word of its own in his Dictionary of Legal Usage (not available from OneLook). Macmillan Dictionary is the only general-purpose dictionary accessible from OneLook in which accusee has an entry, where it is defined as “someone who is accused; a defendant”, with an example dating from 2015: “The defendant or accusee must be subpoened by the small claims court via registered mail paid by the accuser or plaintiff”13. Of course the opinion of the self-appointed guardians of the language has rarely been able to derail the adoption of a word and we can guess there was a similar debate about escaper and escapee, which did not stop the latter noun from winning the contest as of 1936 (NgV).

18 From the 163 examples derived from OneLook general-purpose dictionaries and the various words collected by Bauer (1983), Barker (1998), Plag (2003) and Mühleisen (2010) that we have checked, it appears that only the following have been found to date from the 20th century according to NgV:

19Table 3. -ee nouns dating from the 20th century (NgV)

abortee (1905), appraisee (1926), asylee (1915) auditionee (1933), cohabitee (1928), contactee (1910), counselee (1913), enshrinee (1945), excludee (1925), kidnappee (1910), relocatee (1921), resurrectee (1954), retrainee (1934), stalkee (1915)

20There is little doubt that NgV’s dates of first attestations are as correct as possible since they are based on occurrences from its large corpus of books (cf. section 1, third paragraph), even though it cannot be ruled out that some of these words appeared earlier, whether in spoken language or in books not included in the Google Books project. Some nouns in -ee saw an upswing in frequency as of the 1950s, when the preoccupations of society reflected the need for such terms. Contactee is a good example of this. This noun had a low frequency for nearly four decades in its meaning “a person contacted” and a spectacular takeoff after World War II, when it took on the specific meaning of “a person who claims to have been contacted by creatures from outer space”. As seen above (paragraph below Table 2), franchisee and trainee also increased their frequency significantly when they changed meanings.

21 Attempting to submit more of Mühleisen’s ‘new words’ that she took from the Internet to NgV shows mitigated results. Thus, her 14 “new -ee words” formed with verb bases ending in -cate or -gate”, which are not recorded in general-purpose dictionaries, yield the ensuing data.

22Table 4. Mühleisen’s new -ee nouns formed from verb bases in -cate or -gate (Mühleisen 2010: 23-24)

abrogatee (not found in NgV), confiscatee (NgV, first attestation: 1830), excommunicatee (not found in NgV), indicatee (NgV, first attestation:1803), instigatee (NgV, first attestation: 1828), investigatee (NgV, first attestation:1856), lubricatee (not found in NgV), medicatee (id.), navigatee (id.), placatee (id.), propagatee (id.), relegatee (id.), self-deprecatee (id.), suffocatee (id.).

  • 14 About the relevancy of considering the Internet as a corpus, see the discussions of Kilgarriff & Gr (...)

23The nouns from the table above listed in NgV all entered the English language in the 19th century. Those which do not appear in OneLook dictionaries, OED or NgV may of course be hapax legomena which were collected by Mülheisen from the Internet “from official documents and serious news reports, to semi-private weblogs and chats in internet forums” (Mülheisen 2010: 14, cf. section 3)14.

  • 15 As of today, resurrectee has not gained entry in OneLook dictionaries or in OED.

24 Given that close to 74% of -ee nouns in Mülheisen’s corpus appear in lexical solidarity with -er nouns (applauder / applaudee, etc., cf. paragraph below Table 2), the -ee suffix is definitely morphologically productive as far as ad-hoc and nonce words go, since there are seemingly few limitations on their formation. Some of them, like cheatee or laughee have achieved lexicalisation by being recorded in OED. As far as established nouns are concerned, since only 14 are confirmed by NgV to have been coined in the 20th century15, it entails that the -ee suffix is morphologically productive but not to the extent it is claimed to be by the authors mentioned above.

4. The frequency of well-established nouns suffixed with -ee

25The Collins Dictionary’s recorded usage charts give the peak year of usage of words suffixed with -ee and end in 2008, but this dictionary also contains written sentence examples dating from later years (e.g. 2015, 2016, 2017). Google’s on-line free program Ngram Viewer also gives peak years of usage and ends in 2019. As seen in the sample below, the 163 words of our OneLook corpus were still recorded as used in NgV up to the years 2010-2019, of course with different frequency percentages. The same does not hold true for Collins D.’s usage charts where some items stopped being used in the last quarter of the 20th century, but most formations were still used in the 2010s as indicated in the written examples provided.

26Table 5. Collins D’s and NgV’s frequency of usage

  • 16 One of the few examples where -ee cannot be parsed out ("a person to whom a lease is granted, lat (...)

a. relatively frequent: 0,00XXXXXXXX or 0,000XXXXXXX

employee: one of the most commonly used 4,000 words in Collins D.; the most frequent of all -ee suffixed nouns in NgV, 2019: 0,0030594418%

refugee: one of the most commonly used 10,000 words in Collins D., 2016; NgV, 2019: 0,0006818828%

trainee: one of the most commonly used 10,000 words in Collins D., 2016; NgV, 2019: 0,0001371550%

trustee: one of the most commonly used 10,000 words in Collins D., 2016; NgV, 2018: 0,0004969283%

devotee: one of the most commonly used 30,000 words in Collins D., 2014; NgV, 2019: 0,0000858811%

interviewee: one of the most commonly used 30,000 words in Collins D., 2016; NgV, 2018: 0,0001686282%

lessee16: one of the most commonly used 30,000 words in Collins D., 2008; NgV, 2019: 0,0001283863%

licensee: one of the most commonly used 30,000 words in Collins D., 2006; NgV, 2019: 0,0001612525%

nominee: one of the most commonly used 30,000 words in Collins D., 2016; NgV, 2019: 0,0001220228%

referee: one of the most commonly used 30,000 words in Collins D., 2017; NgV, 2019: 0,0001435769%

(9 items)

b. moderately frequent: 0,0000XXXXXX or 0,00000XXXXX

amputee: Collins D., 2013; NgV, 2019: 0,0000164918%

appointee: Collins D., 2017; NgV, 2018: 0,0000303803%

auditee: Collins D., 2008; NgV, 2019: 0,0000117247%

assignee: Collins D., 2007; NgV, 2018: 0,0000545575%

devotee: Collins D., 2014; NgV, 2019: 0,0000858811%

examinee: Collins D., 2008; NgV, 2018: 0,0000402513%

patentee: Collins D., 2008; NgV, 2018: 0,0000243015%,

transferee: Collins D., 2008; NgV, 2019: 0,0000688841%

abductee: Collins D., 2008; NgV, 2018: 0,0000031333%

internee : Collins D., 2008; NgV, 2019: 0,0000054820%, etc.

(73 items)

c. infrequent: 0,000000XXXX

presentee: Collins D., 2005; NgV, 2019: 0,0000006287%

selectee: Collins D., 2006; NgV, 2019: 0,0000007601%

standee: Collins D., 2003; NgV, 2019: 0,0000005886%

vaccinee: Collins D., 2004; NgV, 2019: 0,0000007025%, etc.

(47 items)

d. rare: 0,00000000XXX (the words below are not recorded in Collins D.)

attackee: NgV, 2019: 0,0000000807%

enshrine: NgV, 2019: 0,0000000399%

infectee: NgV, 2019: 0,0000000783%

petitionee: NgV, 2018: 0,0000000659%, etc.

(34 items)

27As shown above, 82 items in -ee are still relatively to moderately frequent vs. 47 infrequent and 34 rare. In terms of frequency all items suffixed with -ee were still used in 2010-2019, according to NgV, with different degrees, from frequent to rare, and for most of them (when so noted) up to 2008 or the late 2010s in Collins D. Quite a few -ee nouns, which appeared in the 18th and 19th centuries and stopped being used for many years, registered new peaks between the years 1987 and 2000 (e.g. mentee, tutee, presumably because of their usage in the academic world; it should be noted that both have an entry in OneLook general-purpose dictionaries).

5. Stress variation

28One of the basic stress rules of English is that no word is supposed to have two adjacent stresses (Carr 2012: 172173), e.g. ˌderiˈvation (< deˈrive), ˌinspiˈration (< inˈspire), where -ion imposes primary stress one syllable back and, consequently, secondary stress has to be placed at least two syllables before primary stress, here by default on the first syllable, which leads to remetrification. Words contravening this stress-clash rule are: (a) compounds: ˈteenˌager (noun), ˌlong-ˈlegged (adjective), etc.; (b) transparent prefixed formations: ˌreˈmake, ˌunˈthaw (verbs), etc. (Fournier 2010; Carr 2012: 189-190; Trevian 2015: 437438; Dabouis 2019). Exceptions in long words are few, chiefly words formed with electr(o)- (eˌlecˈtricity, eˌlecˈtronic etc.), which all have a variant amenable to the stress rule laid out above (ˌelecˈtricity, ˌelecˈtronic, etc.). As pointed out in the Longman Pronunciation Dictionary (2000, henceforth LPD), there are other optional or compulsory cases in disyllables with final primary stress: (ˌ)arˈcade, (ˌ)arˈcane, (ˌ)bloˈckade, (ˌ)tyˈphoon, ˌbamˈboo, ˌChiˈnese, etc. (cf. paragraph below Table 7).

29 When the appropriate conditions are met, namely when -ee attaches to a base with final stress (in most cases a semantically opaque prefixed verb), the nouns thus formed are apt to allow two consecutive stresses. There are 75 nouns in -ee derivable from a base with final primary or secondary stress in our OneLook corpus. Noted with a penult pattern as its first pronunciation in the English Pronouncing Dictionary (2018) and in LPD, employee ([010] + variant in [201]), the most common of -ee suffixed nouns (cf. Table 5), has been cited by Duchet (1991: 13) as a case of “stress retraction” obeying a general tendency in English. A conflict may be postulated between stress preservation and -ee’s auto-stressing rule resolved in favour of the first pronunciation (emˈployee < emˈploy, on the model of emˈployer). Yet, besides retiree, stressed [201], [021] or [010], there are no other cases of deverbal derivatives in -ee confirming this evolution (Trevian 2015: 77). Furthermore, in US English, employee ([021]) remains apparently the only sanctioned pronunciation.

30 Of the 75 relevant nouns taken from OneLook dictionaries, the following have, or may have, two adjacent stresses.

  • 17 Pronunciation has been verified for each in OED and OneLook dictionaries and, when available ther (...)

31Table 6. Nouns suffixed with -ee allowing two consecutive stresses17

[021] GB and US English

abusee, adoptee, advisee, appointee (+ var. [201] in GB), appraisee, arrestee, attendee, awardee, devisee, distrainee, divorcee, electee, enshrinee, escapee (+ var. [201] in GB), evictee, expiree, exploitee, indorsee (+ var. [201] in GB), infectee, promotee, rejectee (+ var. [201] in GB), relocatee (here the prefixed form is transparent [2031] + var. [2301] and [2001]), resignee, retiree (+ var. [201] or [010] in GB), retrainee (another case of transparent prefixed form [231]), seducee, selectee + (not derived from a verb with a prefix): asylee [021] < asylum [010], parolee [021] < parole [01].

[021] US English

abductee, abortee, accusee (not listed in British dictionaries), addressee, allottee (+ var. [201]), assignee, attackee (not listed in British dictionaries), confinee, consignee (+ var. [201]), consultee, debauchee, deportee, detainee (+ var. [201]), devotee, dilutee (+ var. [201]), dischargee, employee, enlistee, enrollee (+ var. [201]), expellee, importee (not listed in British dictionaries), indictee, inductee (+ var. [201]), internee, invitee (+ var. [201]), obligee (+ var. [201]), permittee (+ var. [201]), presentee (+ var. [201]), reservee; the following nouns are derived from a verb stressed [(-)102] in US English: dedicatee ([2031] + var. [2001]), educatee ([2031] + var. [2001]), interrogatee ([02031] + var. [02001]), interviewee ([2031]), supervisee ([2031]).

  • 18 Even the spelling variant conferree which is given in several OneLook dictionaries, including Col (...)
  • 19 Whereas the preˌsenˈtee pronunciation would result from the verb preˈsent (senses: “a person pres (...)

3263 items impose or allow two stresses in a row in American English. 29 of these nouns also produce [021] in British English. Whereas there is a variant with stress clash in dedicatee, educatee, interrogatee, interviewee and supervisee in American English, no adjacent stresses are noted in communicatee and delegatee (allocatee is only recorded in Merriam Webster’s D., but with no transcription). The same goes for representee in Merriam Webster’s D (listed but no transcription given) but not for committee, in the obsolete sense of “a person to whom a trust or charge is committed”, which Merriam Webster’s D. notes with a [201] stress pattern. Confirmee and excludee are not listed in any recent American dictionary giving transcriptions. The US pronunciation of ˌtransfeˈree and ˌtransporˈtee, which is identical to British English, merely reflects the fact that, in American English, these two verbs are generally stressed initially: ˈtransfer and ˈtransport. Once the preceding cases have been discarded, the only nouns in -ee derivable from an unparsable two-syllable verb with a prefix which do not sanction two successive stresses in American English are: absentee [201], conferee (id.)18 and referee (id.). Absentee is probably interpreted as derived from the adjective (ˈabsent), like presentee, in the sense of “someone who is present”,19 despite the etymology provided in OED: “< absent (verb) + -ee suffix, originally after Anglo-Norman abscenté. Conferee and referee may result from adjacent motivation on the models of ˈconference < conˈfer and ˈreference < reˈfer.

33 Of all the nouns suffixed with -ee susceptible to stress clash, communicatee, delegatee, committee (suffixed form), conferee and referee seem to be the only ones whose stress patterns remain the same in British English and American English (respectively [02001], [2001] and [201]).

34 Contrary to -ee, the stress-bearing -ese suffix entails remetrification, with no possible variant, when added to a base with final stress whose first syllable is reduced in [ə], [ɪ] or [u], cp. Ceyˈlon > ˌCeyloˈnese with aˈward > aˌwarˈdee.

35Table 7. Remetrification of words suffixed with -ese

Ceyˈlon > ˌCeyloˈnese, Jaˈpan > ˌJapaˈnese, Miˈlan > ˌMilaˈnese, Suˈdan > ˌSudaˈnese, etc.

36However, two-syllable suffixed words in -ese are not immune to adjacent stresses as shown by Chinese, Maltese, Bernese, Burmese, which are all noted [21] in LPD. According to Wells (LPD: 742), “stress shift can apply to any word that has a secondary stress before its primary stress” (the example of ˌantique ˈchair is provided in Wells’s discussion), henceˌChinese ˈrestaurant, ˌMaltese ˈfalcon, ˌBernese ˈOberland, ˌBurmese ˈmountains where the suffixed forms in -ese are attributive adjectives. In a similar fashion, two-syllable nouns noted [21] in LPD will sanction weak preservation (about this concept, cf. Burzio, 1994: chapter 6) when affixed with -eer (another stress-bearing suffix): (ˌ)harˈpoon > ˌharpooˈneer, ˌrouˈtine > ˌroutiˈneer, etc., contrary to gaˈzette (where the rhyme of the first syllable is [ə]) and thus entails remetrification) > ˌgazetˈteer. Two-syllable nouns containing the autostressed suffix -ette also allow adjacent stressing. Thus, ladette and punkette are noted with a double stress ([21]) in Collins D. and Oxford Dictionaries whilst other items, such as rockette and Smurfette, which are recorded with no phonetic transcription in OneLook neologism or slang dictionaries, can also have two consecutive stresses, as evidenced by audio and video recordings (e.g. the Smurf cartoons: https://www.youtube.com/​watch?v=RgTpFaAm2f4).

37 It is well known that stress clash is avoided in noun phrases with words having primary stress on their last syllable: She’s a ˌJapanese (attributive adjective) ˈgirl vs. She’s ˌJapaˈnese (predicative adjective), and similarly ˌChinese ˈlanterns vs. they’re ˌChiˈnese, ˌfourteen ˈpeople vs. they were ˌfourˈteen, ˌrefugee ˈcamps vs.ˌmany ˌrefuˈgees, etc. Whereas not all two-syllable words susceptible to stress shift have, at least potentially, a secondary stress (e.g. carˈtoon, LPD but ˌcartoon ˈartist, ˌcartoon ˈcharacter, etc.) as Wells implies with antique, the key factor is a non-reduced first vowel since words like baroque or maroon which have a [ə] in the rhyme of their first syllable cannot shift stress (Baˌroque ˈmusic, a maˌroon ˈsweater).

38 Dabouis (2019) has attempted to account for stress preservation overriding clash. His study is focused on transparently suffixed words (thus excluding LPD-sourced unparsable two-syllable words in [21] such as ˌarˈcane, (ˌ)harˈpoon, ˌrouˈtine, etc.) and scrutinises the models of “fake cyclicity” (Collie 2007, 2008), “dual route of lexical access” (stress preservation hinging on whether the base is more frequent than its derivative, Hay 2001, 2003) and “weighted constraints” (Pater 2009, 2016). However, when dealing with -ee, Dabouis (2019) comes across the same discrepancy between -ee and -ese: “-ee also imposes final accent, which is marked in English, especially for nouns. This accentual property of the suffix may facilitate its parsing and therefore preservation from the base. However, one would have to explain the behaviour of the words containing -ese.”

39 To account for the violation of the stress-clash constraint in nouns in -ee, namely of the kind [+ reduced syl.] + [21] (asylum > aˌsyˈlee, aˈward > aˌwarˈdee, etc.), which does not occur with the suffixes -ese (Ceyˈlon > ˌCeyloˈnese, Jaˈpan > ˌJapaˈnese, etc.) or -eer (gaˈzette > ˌgazetˈteer), it may be posited as Dabouis does (2019) that the -ee suffix is “more decomposable” than others. According to Hay (2003: 137), despite equivalent base-derivative frequency ratios, grayish is more decomposable than scenic because, contrary to words suffixed with -ic, those suffixed with -ish are often less frequent than their bases. Dabouis (2019) notes that the “derivatives which are less frequent than their base and have an open second syllable but nonetheless preserve an accent on the second syllable are all -ee derivatives (debauchee, detainee, remittee) and that could be accounted for if -ee itself turned out to be highly decomposable”. Hay & Baayen (2003) report that -ee is more parsable than other suffixes. Dabouis’s analysis, based on the findings of Hay (2003) and Hay and Baayen (2003), has the merit of offering an explanation as to why suffixed words in -ee sanction weak preservation and thus binary feet one after the other in polysyllabic nouns (apˈpraise > apˌpraiˈsee, etc.), a better proposition than putting this down to mere idiosyncrasies.

6. Conclusion

40In this paper it has been shown that far fewer established words (14 in all) than believed by Bauer, Barker, Plag or Mühleisen are coinages from the 20th century. Most of the examples they give are nonce words or ad-hoc formations, some of them being potentially lexicalisable, or what they call ‘neologisms’ which are often not recorded in NgV’s frequency charts. However, the 163 words recorded in OneLook general-purpose dictionaries and the additional words provided by these authors that I have been able to check from NgV are still used, at least in written works. It can be concluded from formations listed in general-purpose dictionaries and their dates of first attestation given in NgV that the -ee suffix is, morphologically, relatively instead of highly productive.

41 From NgV charts, it appears that the -ee nouns which are listed in them are still used, with various degrees of frequency.

42 Regarding adjacent stresses of the apˌpraiˈsee type, they are nearly always sanctioned in American English and attested in more than one-third of them in British English. It can be advanced that the prevalence of weak preservation over clash stems from the -ee suffix being highly decomposable, which is seemingly not the case with other stress-bearing suffixes such as -ese or -eer. In this respect, it will be interesting in the coming years to check whether more three-syllable nouns suffixed with -ee will convert to [021] in British English.

Top of page

Bibliography

Abasq, Véronique, Quentin Dabouis, Jean-Michel Fournier and Isabelle Girard. The Core of the English Lexicon: Stress and Graphophonology.” Anglophonia French Journal of English Linguistics. 27 (2019); <https://journals.openedition.org/anglophonia/2317>.

Anderson, Stephen R. A-Morphous Morphology. Cambridge: Cambridge University Press, 1992.

Baayen, Harald. “Quantitative aspects of morphological productivity.” In Geert E. Booij and Jaap van Marle, eds. Yearbook of morphology 1992. Dordrecht: Kluwer, 1992: 109-149.

Baayen, Harald. “On frequency, transparency and productivity.” In Geert E. Booij and Jaap van Marle, eds. Yearbook of morphology 1992. Dordrecht: Kluwer, 1993: 181-208.

Baayen, Harald and Rochelle Lieber. “Productivity and English derivation: a corpus-based study.” Linguistics 29 (1991): 801-843.

Baayen, Harald and Antoinette Renouf. “Chronicling the Times: productive lexical innovations in an English newspaper.” Language 72 (1996): 69-96.

Barker, Chris. “Episodic ee in English: a thematic role constraint on word formation.” Language 74 (1998): 695-727.

Barnhart, Robert K., Sol Steinmetz and Clarence L. Barnhart. Third Barnhart Dictionary of New English. New York: Wilson, 1990.

Bauer, Laurie. English Word-Formation. Cambridge: Cambridge Textbooks in Linguistics, 1983.

Bauer, Laurie, Rochelle Lieber and Ingo Plag. The Oxford Reference Guide to English Morphology. Oxford: Oxford University Press, 2013.

Burzio, Luigi: Principles of English Stress. Cambridge: Cambridge University Press, 1994.

Carr, Philip: English Phonetics and Phonology. An Introduction. 2nd Edition. Chichester: Wiley Blackwell, 2012.

Collie, Sarah. English Stress Preservation and Stratal Optimality Theory. Ph.D. dissertation. Edinburgh: University of Edinburgh, 2007.

Collie, Sarah. “English Stress Preservation: the Case for ‘Fake Cyclicity’.” English Language and Linguistics 12/3 (2008): 505–53.

Dabouis, Quentin. “When Accent Preservation Leads to Clash.” English Language and Linguistics 23/2 (2019): 363-404.

Dixon, R.M.W. Making New Words: Morphological Derivation in English. Oxford: Oxford University Press, 2014.

Duchet, Jean-Louis. Code de l’anglais oral. 1st edition. Paris: Ophrys, 1991.

Fournier, Jean-Michel. Manuel d’anglais oral. Paris: Ophrys, 2010.

Garner, Bryan A. Garner’s Dictionary of Legal Usage. 3rd edition. Oxford: Oxford University Press, 2011 [1987].

“Google n-grams and pre-modern Chinese” (no author mentioned) Digital Sinology, Modern methods in early Chinese studies (2015); <https://digitalsinology.org./google-ngrams-pre-modern-chinese/>.

Guierre, Lionel. Drills in English Stress Patterns. 4th edition. Paris: Armand Colin-Longman, 1984 [1968].

Hay, Jennifer. 2001. “Lexical Frequency in Morphology: Is Everything Relative?” Linguistics 28/6 (2001): 1041–70.

Hay, Jennifer. Causes and Consequences of Word Structure. London: Routledge. 2003.

Hay, Jennifer and Harald Baayen. “Phonotactics, Parsing and Productivity.” Italian Journal of Linguistics 15/1 (2003): 99–130.

Huddleston, Rodney and Geoffrey K. Pullum. The Cambridge Grammar of the English Language. Cambridge: Cambridge University Press, 2002.

Jones, Daniel. English Pronouncing Dictionary. 19th edition. Peter Roach, Jane Setter and John Esling, eds. Cambridge: Cambridge University Press, 2018 [1917].

Kies, Daniel. Corpus research and your presentations. 2020: <https://rhetory.com/corpus-research/#/>.

Kilgarriff, Adam and Gregory Grefenstette. “Introduction to the special issue on the web as corpus.” Computational Linguistics 29 (2003): 333–347.

Mühleisen, Susanne. Heterogeneity in Word-Formation Patterns. Amsterdam/Philadelphia: John Benjamins Publishing Company, 2010.

Oxford English Dictionary, <www.oed.com>.

Pater, Joe. “Weighted Constraints in Generative Linguistics.” Cognitive Science 33/6 (2009): 999–1035.

Pater, Joe. “Universal Grammar with Weighted Constraints.” In John J. McCarthy and Joe Pater, eds. Harmonic Grammar and Harmonic Serialism. London: Equinox (2016): 1–35.

Plag, Ingo. Word-Formation in English. Cambridge: Cambridge University Press, 2003.

Resnik, Philip and Noah A. Smith. “The web as a parallel corpus.” Computational Linguistics 29 (2003): 349–380.

Trevian, Ives. English suffixes, stress-assignment properties, productivity, selection and combinatorial processes. Linguistic Insights Studies in Language Communication. Bern: Peter Lang, 2015.

Wells, John C. Longman Pronunciation Dictionary, 2nd edition. London: Longman, 2000 [1990].

Online general-purpose dictionaries accessible from OneLook <www.onelook. com>

American Heritage Dictionary of the English Language, <http://education.yahoo.com/ reference/dictionary>.

Cambridge Advanced Learner’s Dictionary, <http://dictionary.cambridge.org>.

Collins English Dictionary, with recorded usage charts, <http://www.collinsdictionary.com>.

Dictionary.com; <http://dictionary.reference.com>.

Merriam Webster’s Online Dictionary, 11th Edition, <http://www.merriam-webster.com>.

Macmillan Dictionary, <https://www.macmillandictionary.com>.

Oxford Dictionaries, <https://www.lexico.com>.

Online word frequency search engine

Google Books Ngram Viewer, <https//:books.google.com/ngrams>.

Top of page

Notes

1 For non-sapient nouns like bootee = short woolen socks that babies wear instead of shoes” or goatee = “short pointed beard that covers a man's chin but not his cheeks”, cf. section 2.

2 In his well-received book, Dixon (2014: 51) curiously states that the -ee suffix is attested in about twenty words.

3 But much less so for other languages, notably Chinese. “Google n-grams and pre-modern Chinese” (no author mentioned, 2015) Digital Sinology, Modern methods in early Chinese studies; https://digitalsinology.org/google-ngrams-pre-modern-chinese/.

4 Even though they can be interesting for linguistic analysis, notably for the identification of nonce words, the crowdsourced dictionary of slang Urban Dictionary, or Wiktionary which is written collaboratively by volunteers, have not been included in this study.

5 Plag (2003: 88), who does not adhere to Guierrian generalisations (e.g. -VʹVʹ(C)), simply describes the suffix -ee as autostressed.

6 Contrary to Guierre (1984), Duchet (1991) and Abasq et al (2019), -aaC has been discarded as the few relevant words are loans or proper nouns, from Dutch or Asian languages, where no pattern of regularity is found: 'advocaat, 'Balaam, 'Canaan, 'kursaal, vs. ˌBalma'caan, Ba'taan, ba'zaar, ˌmeshu'gaas, sa'laam, Trans'vaal.

7 Other OneLook dictionaries may differ in the number of their definitions but they all set apart the diminutive sense of the -ee suffix.

8 The dates of first written occurrences have been taken from NgV.

9 Despite the etymology given in dictionaries, it cannot be excluded that biographee actually resulted from the contrasted pair biographer / biographee.

10 Although synchronically denominal refugee (< refuge) is apparently an anglicisation of refugié which occurred with the arrival in England of many French Protestants following the revocation of the Edict of Nantes (OED).

11 As quite a few nouns from her list, like many in Mülheisen’s corpus, are absent from OneLook dictionaries, OED and NgV (e.g. constipatee, devastatee, remuneratee), I have not been able to check their dates of first written use.

12 As said above, they resorted in most cases to OED for dates of first written attestations.

13 Accusee is listed as having existed since 1803 in NgV.

14 About the relevancy of considering the Internet as a corpus, see the discussions of Kilgarriff & Grefenstette 2003; Resnik & Smith 2003.

15 As of today, resurrectee has not gained entry in OneLook dictionaries or in OED.

16 One of the few examples where -ee cannot be parsed out ("a person to whom a lease is granted, late 15c., via Anglo-French lesee, Collins D.).

17 Pronunciation has been verified for each in OED and OneLook dictionaries and, when available therein (e.g. OED, Collins D., Merriam-Webster’s D., Oxford dictionaries) from audio recordings.

18 Even the spelling variant conferree which is given in several OneLook dictionaries, including Collins D., does not allow the [021] pronunciation.

19 Whereas the preˌsenˈtee pronunciation would result from the verb preˈsent (senses: “a person presented or nominated” or “a person to whom something is presented”).

Top of page

References

Electronic reference

Ives Trevian, The suffix -ee: history, productivity, frequency and violation of stress rulesAnglophonia [Online], 30 | 2020, Online since 20 December 2020, connection on 22 January 2022. URL: http://journals.openedition.org/anglophonia/3504; DOI: https://doi.org/10.4000/anglophonia.3504

Top of page

About the author

Ives Trevian

Université de Paris, UFR d’études anglophones
trevian@univ-paris-diderot.fr

Top of page

Copyright

Licence Creative Commons
Anglophonia – French Journal of English Linguistics est mis à disposition selon les termes de la licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Top of page
Search OpenEdition Search

You will be redirected to OpenEdition Search