Rebracketing (also known asresegmentation ormetanalysis) is a process inhistorical linguistics where a word originally derived from one set ofmorphemes is broken down or bracketed into a different set. For example,hamburger, originally fromHamburg+er, has been rebracketed intoham+burger, andburger was later reused as aproductive morpheme in coinages such ascheeseburger. It is usually a form offolk etymology, or may seem to be the result of validmorphological processes.
Rebracketing often focuses on highly probable word boundaries: "a noodle" might become "an oodle", since "an oodle" sounds just as grammatically correct as "a noodle", and likewise "an eagle" might become "a neagle", but "the bowl" would not become "th ebowl" and "a kite" would not become "ak ite".
Technically, bracketing is the process of breaking an utterance into its constituent parts. The term is akin toparsing for larger sentences, but it is normally restricted to morphological processes at the sublexical level, i.e. within the particular word orlexeme. For example, the worduneventful is conventionally bracketed as [un+[event+ful]], and the bracketing [[un+event]+ful] leads to completely differentsemantics. Rebracketing is the process of seeing the same word as a different morphological decomposition, especially where the newetymology becomes the conventional norm. The namefalse splitting, also calledmisdivision, in particular is often reserved for the case where two words mix but still remain two words (as in the "noodle" and "eagle" examples above).
The namejuncture loss may be specially deployed to refer to the case of an article and a noun fusing (such as if "the jar" were to become "(the) thejar" or "an apple" were to become "(an) anapple"). Loss ofjuncture is especially common in the cases ofloanwords and loan phrases in which the recipient language's speakers at the time of the word's introduction did not realize an article to be already present (e.g. numerous Arabic-derived words beginning 'al-' ('the'), including "algorithm", "alcohol", "alchemy", etc.). Especially in the case of loan phrases, juncture loss may be recognized as substandard even when widespread; e.g. "thehoi polloi", where Greekhoi = "the".
As a statistical change within a language within any century, rebracketing is a very weak statistical phenomenon. Even during phonetic template shifts, it is at best only probable that 0.1% of the vocabulary may be rebracketed in any given century.
Rebracketing is part of the process oflanguage change, and often operates together withsound changes that facilitate the new etymology.
Rebracketing is sometimes used forjocular purposes, for examplepsychotherapist can be rebracketed jocularly asPsycho the rapist, andtogether in trouble can be rebracketed jocularly asto get her in trouble.[1]
Before the increased standardization of the English language in the modern period, many new words entered its lexicon in exactly the way just described. A 15th century English cook may once have said something like: "Ah, I found this ewt and this nadder in my napron while baking numble-pie." A few generations later the cook's descendant would have said: "Ah, I found this newt and this adder in my apron while baking (h)umble-pie." Over the course of time these words were misheard and resegmented:ewt becamenewt,nadder becameadder,napron becameapron,numble-pie became(h)umble pie. The force behind these particular resegmentations, and by far the most powerful force behind any such resegmentations in the English language, was the "movable-n" of the indefinite articlea(n), of the possessive pronounsmy(n) andthy(n), and of the old dative case of the definite articlethe(n). The biformsno/none, the prepositionsin andon, the conditional conjunctionan even, the shortened formn (and), and the inflectional endings in-n may also have played a part. Through the process ofprothesis, in which the sound at the end of a word is transferred to the beginning of the word following, or converselyapheresis, in which the sound at the beginning of a word is transferred to the end of the word preceding, old words were resegmented and new words formed. So through prothesisan ewt becamea newt. Conversely through aphaeresisa nadder becamean adder,a napron becamean apron, anda numble-pie becamean (h)umble-pie. Many other words in the English language owe their existence to just this type of resegmentation: e.g.,nickname,ninny,namby-pamby,nidiot/nidget,nonce word,nother, andnotch through prothesis ofn;auger,umpire,orange,eyas,atomy,emony,ouch, andaitch-bone, through aphaeresis ofn.[2]
Many productive affixes have been created by rebracketing, such as-athon fromMarathon,-holic fromalcoholic, and so on. These unetymological affixes arelibfixes.
The origins of the wordhamburger were in a form of ground meat dish originating fromHamburg,Germany. The bracketing of the original washamburg‧er, but after its introduction into theUnited States, it was soon factorized asham‧burger (helped byham being a form of meat). This led to the creation of the independent suffix-burger: chickenburger, fishburger, etc. In the original etymology,burg was town andburger was a resident, or something related to the town; after refactorization it becomes a chunk of meat for a sandwich, although a hamburger does not contain ham.[3]
The English wordoutrage is aloanword from French, where it was formed by combining the adverboutre (meaning "beyond") with the suffix-age, rendering a bracketing ofoutr‧age and a meaning of "beyondness" (from what is acceptable). The rebracketing as a compound ofout- with the noun or verbrage has led to both a different pronunciation than the one to be expected for such a loanword (compareumbrage) and an additional meaning of "angry reaction" not present in French.
The Englishhelico‧pter, fromGreekheliko- ("turning") andpteron ("wing"), has been rebracketed to modernheli‧copter (as injetcopter,heliport).
cybern‧etics (from Greekkubernān and-ētēs) has been split intocyber‧netics (as incyber‧space).
prosthodontics contains the prefixprosth(o)-, which arose by misdivision ofprosthe‧tic intoprosth- and-etic. The word comes from Greekpros ("in front of") andthē-, theroot of the verbtithēmi ("I place").
The dog breedLabrad‧oodle (a cross between aLabrador Retriever andPoodle) has been rebracketed toLabra‧doodle, leading to the-doodle suffix in other Poodle crossbreeds such as theGoldendoodle and Aussiedoodle.
The wordalcohol‧ic derives fromalcohol (itself a junctureless rebracketing of Arabical-kuḥl) and-ic. Words for other addictions have formed by treating-holic as a suffix:workaholic,chocoholic, etc.
In Romance languages, repeated rebracketing can change an initiall to ann (first removing thel by analyzing it as the definite articlel', and then addingn by rebracketing from the indefinite articleun), or the reverse. Examples include:[4]
Latin *libellu ("level") becomingnivel in Portuguese, Romanian, and Spanish, andniveau in French.
Latinunicornis ("unicorn") becamelicorne in French, viaunicorne >une icorne (a unicorn), and finally, with juncture loss,l'icorne (the unicorn) >licorne.
InSwahili,kitabu ("book") is derived fromArabickitāb (كتاب). However, the word is split as a native Swahili word (ki- +tabu) and declined accordingly (pluralvitabu).[5] This violates the originaltriliteral root of the original Arabic (K-T-B).
Many words coined in a scientific context as neologisms are formed with suffixes arising from rebracketing existing terms. One example is the suffix-ol used to namealcohols, such asmethanol. Its origin is the rebracketing ofal‧cohol asalcoh‧ol. The wordalcohol derives from the Arabical-kuḥl, in whichal is the definite article andkuḥl (i.e.,kohl) is based on the Semitic triliteral rootK-Ḥ-L.[6] The suffix-ome, as ingenome, is occasionally suggested as being a rebracketing ofchromo‧some aschromos‧ome,[7] but see discussion atOmics asserting a derivation from other, similar coinages.
InScottish Gaelic, thedefinite article is pronounced run together with vowel-initial nouns without audible gap, similar toFrench. This union has provided a rich source of opportunities for rebracketing. Historically the article's various case-, number-, and gender-specific forms ended in either a vowel, a nasal or an/s/, the latter later becoming an/h/ over time. Over time, the last syllable of the article was either eroded completely or weakened and partially lost, but where rebracketing had occurred, what had been the final consonant of the article came to be treated as the initial of the following noun. Example:an inghnean ( <*(s)indā inigenā) gave rise to an alternative forman ighean (the girl) this in turn becomingan nighean. As a second, more extreme example, the Scottish Gaelic words fornettle includeneanntag, eanntag, deanntag, and evenfeanntag. In addition, many forms of the article cause grammatically conditionedinitial consonant mutation of the following noun. The original cause of this mutation in theCeltic languages was an across-the-boardchange of pronunciation of certain non-geminate consonants where they were either trapped between two vowels, or else between a vowel or certain other consonants. Mutation gave rise to yet more possibilities for reanalysis, the formfeanntag mentioned earlier possibly being one such example. Calder 'A Gaelic Grammar' (1923) has a useful list.
As demonstrated in the examples above, the primary reason of juncture loss in English is the confusion between "a" and "an". InMedieval script, words were often written soclose together that for someMiddle English scholars it was hard to tell where one began and another ended. The results include the following words in English:
adder: Middle Englisha naddre ("a snake") taken foran addre.
aitchbone: Middle Englisha nachebon ("a buttock bone") taken foran hach boon.
another, formed by combining "an other" into one word, is sometimes colloquially split into "a nother" and a qualifier inserted as in "a whole nother issue".
chord: Middle Englishaccord (harmony) taken fora cord, later influenced by "chord" (archaic name for a string), which has another etymology.
decoy: Most commonly thought to stem from Dutchde kooi, in whichde is the definite article andkooi means cage. An alternative theory is that the Dutch compound nouneendenkooi, earlier spelledeendekooi, meaning "duck decoy", fromeend "duck" +kooi, was reanalyzed and split, in the process of being transferred to English, aseen dekooi, in whicheen is the Dutch indefinite article.
eyas: Middle Englisha niyas (from Frenchniais from Late Latinnidiscus (from Latinnidus = "nest")) taken foran eias.
humble pie: Middle Englisha numble taken foran umble (ultimately from Latinlumbulus, this is also an example ofhomorganicness).
lone: Middle Englishal one (all one) taken fora-lone.
newt: Middle Englishan eute (cognate witheft) taken fora neute.
nickname: Middle Englishan eke name ("an additional name") taken fora neke name.
the nonce: Middle English, for old Englishþen ānes (the one [occasion]).
nuncle (dialectal form ofuncle): Middle Englishmine uncle taken formy nuncle.
omelette: Seventeenth-century English loanword from French, developed there via earlier formsamelette,alemette andalemelle fromla lemelle ("the omelette") taken forl'alemelle; ultimately from Latinlamella ("blade"), perhaps because of the thin shape of the omelette (SOED).
ought ["zero"]: Middle Englisha nought ("a nothing") taken foran ought. Ultimately distinct from Old Englishnaught ("nothing"), of complex and convergent etymology, fromna ("not") andwight ("living thing, man"), but cf.aught ("anything", "worthy", etc.), itself ultimately fromaye ("ever") andwight (SOED).
synth: short forsynthesizer and evoked sometimes as an abbreviation of "synthetic" – from Greek-derived affixes "syn-" (together) and "thesis" (put) – literally "put together" (compare Latin-derived compose)
tother: Middle English (now dialectal)that other taken forthe tother.
umpire: Middle Englisha noumpere taken foran oumpere.
Old Frenchlonce ("lynx") taken forl'once, thus giving rise toonce (hence English:ounce), now more often applied to thesnow leopard.
Old Frenchune norenge ('an orange') taken forune orenge.[8]
boutique from Greek-derived Latinapotheca, a change found in some Romance languages (e.g. Italianbottega, Spanishbodega, Sicilianputìa), a putative proto-Romancel'aboteca orl'abodega taken forla +lexeme.
licorne ("unicorn") from rebracketing ofl'icorne;icorne itself comes from rebracketing of Old Frenchunicorne asune icorne.
lierre ("ivy") from Old Frenchliere, a rebracketing ofl'iere.
Lille (a city in French Flanders) from rebracketing ofl'isle ("the island")
Dutch shares several examples with English, but also has some of its own. Many examples were created by reanalysing an initialn- as part of a preceding article or case ending.
adder: As in English.
arreslee (horse-drawn sleigh): From early modern Dutcheen (n)arreslede, fromnar "fool, jester" +slede "sleigh".
avegaar "auger": As in English.
omelet "omelette": As in English.
spijt "pity, regret": From Middle Dutchdespijt, from Old Frenchdespit "spite". Reanalysed asde spijt "the pity".
Rijsel "Lille" : fromter IJsel "at the Isle", reanalyzed aste Rijsel "at Lille".
Perhaps the most common case of juncture loss in English comes from the Arabical- (mentioned above), mostly via Spanish, Portuguese, and Medieval Latin:
Arabicar-rub in Spanish asarroba, a unit of measure.
Arabicaz-zahr ("the dice") in Spanish asazar, "randomness", and in French and English as "hazard"
Arabical-fīl ("the elephant") in Spanish asalfil "chess bishop" and in Italian asalfiere "chess bishop"(whose Russian nameслон (slon) also means "elephant").
Junctural metanalysis played a role in the development of new words in the earliest period of Greek literature: during the oral transmission of the Homeric epics. Many words in the Homeric epics that are etymologically inexplicable through normal linguistic analysis begin to make some sense when junctural metanalysis at some stage in the transmission is assumed: e.g., the formulaeche nedumos hypnos "sweet sleep held (him)" appears to be a resegmentation ofechen edumos hypnos. Steve Reece has discovered several dozen similar instances of metanalysis in Homer, thereby shedding new light on their etymologies.[10]
Juncture loss is common in later Greek as well, especially in place names, or in borrowings of Greek names in Italian and Turkish, where particles (εις, στην, στον, σε) are fused with the original name.[11][12][13] In theCretan dialect, these- prefix was also found in common nouns, such assecambo ortsecambo <se- +cambo 'a plain'.[14]
^For examples of resegmentation in Middle English in various phonetic environments, see Steve Reece,Junctural Metanalysis in Middle English, in Reece, Steve,Homer's Winged Words (Leiden: Brill, 2009) 15-26. Also Reece, Steve, "Some Homeric Etymologies in the Light of Oral-Formulaic Theory," Classical World 93.2 (1999) 185-199.https://www.academia.edu/30641357/Some_Homeric_Etymologies_in_the_Light_of_Oral-Formulaic_Theory
^Pierre, Alexandre (1983). "Langue arabe et kiswahili" [Arabic and Kiswahili].Langue arabe et langues africaines [Arabic and African languages] (in French). Conseil international de la langue française. pp. 9–10.ISBN9782853191258.ainsi kitabu كتاب "livre" est interprété /ki-tabu/ avec pluriel /vi-tabu/.
^Reece, Steve (2009).Homer's Winged Words: The Evolution of Early Greek Epic Diction in the Light of Oral Theory. Leiden and Boston: Brill.ISBN978-90-04-17441-2. Also, Reece, Steve, "Some Homeric Etymologies in the Light of Oral-Formulaic Theory,"Classical World 93.2 (1999) 185-199.Some Homeric Etymologies in the Light of Oral-Formulaic Theory
^abBourne, Edward G. (1887). "The Derivation of Stamboul".American Journal of Philology.8 (1). The Johns Hopkins University Press:78–82.doi:10.2307/287478.JSTOR287478.
^Marek Stachowski, Robert Woodhouse, "The Etymology of İstanbul: Making Optimal Use of the Evidence"Studia Etymologica Cracoviensia20: 221–245 (2015)doi:10.4467/20843836SE.15.015.2801
^abcC. Desimoni, V. Belgrano, eds., "Atlante Idrografico del Medio Evo posseduto dal Prof. Tammar Luxoro, Pubblicata a Fac-Simile ed Annotato",Atti della Società Ligure di Storia Patria, Genoa, 18675:103cf.Luxoro Atlas
Hendrickson, Robert.QPB Encyclopedia of Word and Phrase Origins. New York: Facts on File, Inc., 1998.
Reece, Steve.Homer's Winged Words: The Evolution of Early Greek Epic Diction in the Light of Oral Theory. Leiden and Boston: Brill, 2009. [This book is concerned primarily with junctural metanalysis in ancient Greek, but it includes a chapter on Middle English, and it catalogues examples in many other languages: Sanskrit, Tocharian, Old Church Slavic, Latin, Frankish, Venetian, Turkish, Italian, French, Spanish, Haitian, German, Dutch, Irish, Gaelic, Welsh, and Arabic.]
Pickett, Joseph P.The American Heritage dictionary of the English language.—4th ed. New York: Houghton Mifflin Co., 2000. [also: * Morris, William.The American Heritage dictionary of the English language.—new college ed. Boston: Houghton Mifflin Co., 1976. [also: "New College Ed.", ed. William Morris. 1976]
Vizetelly, Frank H.Funk & Wagnalls Practical Standard Dictionary of the English Language New York: Funk & Wagnalls Co., 1931.
Webster, Noah.American Dictionary of the English Language. New Haven: S. Converse, 1828.
