[in progress] Chapter 14: Historical Linguistics
In many languages, root morphemes may combine with different inflectional affixes (see Section 5.2 for discussion of root morphemes and Section 5.7 for discussion of inflectional affixes). The resulting set of words is called the root’s inflectional paradigm or just paradigm for short. Verb paradigms are sometimes called conjugations, while noun and adjective paradigms are sometimes called declensions, especially in language learning courses. An example paradigm is given in Table 14.1 for the English verb move.
|move||present tense, except for third-person singular (I move)|
|moves||third-person singular present (she moves)|
|moved||past tense and past participle (I moved, I have moved)|
|moving||present participle (I am moving)|
Many other English verbs (check, end, sign, etc.) have this same basic pattern in their paradigms, so it can also be useful to talk about that pattern as a kind of paradigm itself, separate from specific verb roots. This basic verb paradigm in English is given in Table 14.2.
|ROOT||present tense, except for third-person singular|
|ROOT-s||third-person singular present|
|ROOT-ed||past tense and past participle|
There are other verb paradigms in English with different patterns. Some verbs have different past tense and past participle forms, such as show, which has the paradigm in Table 14.3.
|show||present tense except for third-person singular (I show)|
|shows||third-person singular present (she shows)|
|showed||past tense (I showed)|
|shown||past participle (I have shown)|
|showing||present participle (I am showing)|
In addition, many verbs undergo changes in pronunciation in some parts of the paradigm, such as internal change (changing one or more phones in the root) or even suppletion (changing the entire root); see Section 5.3 for discussion. Some examples of paradigms with internal changes and suppletion are given in Table 14.4, with internal changes highlighted in blue and suppletion highlighted in orange. Note that gone has an internal change, with the vowel of the root changing from [o] to [ɒ], thouɡh this is not apparent from the spelling.
|ring||go||present tense except for third-person singular|
|rings||goes||third-person singular present|
English has many different verb paradigms, mostly with internal change. In addition to the ring-rang-rung pattern, there are patterns like fly-flew-flown, ride-rode-ridden, speed-sped-sped, speak-spoke-spoken, and think-thought-thought. There are even degenerate patterns like hit-hit-hit, put-put-put, and quit-quit-quit, which have only a single form where many other verbs have three different forms.
Differences between paradigms can be a source of language change. Languages often shift to make their paradigms more consistent, either within a single paradigm or across multiple paradigms. They may copy existing patterns from elsewhere in the same paradigm or from a completely different paradigm. This kind of copying of paradigm patterns is part of a general type of language change called analogy. Analogy was introduced in Section 14.3 for sound change, but it is also a significant driving force in morphological change.
Recall from Section 14.3 the following sound change from Old English to Middle English that explains why words like sword are pronounced with [s] rather than [sw]:
- [sw] > [s] / ▁ back vowel
For the verb swear, the root and most other forms have a front vowel, so these forms did not change [sw], which still remains in Modern English: swear, swear, and swearing. However, swear had the same paradigm as speak, so it had a back vowel in the past tense and past participle forms: swore and sworn, just like spoke and spoken. If the [sw] > [s] change had affected swore and sworn as expected, then we would now pronounce them with [s], just like we do for sword. However, by analogy with the rest of the swear paradigm, the [sw] in all of swear’s forms was retained, blocking the sound change from applying to swore and sworn, which keeps the paradigm more consistent.
Analogy commonly results in a reduction in the total number of allomorphs a root has in its paradigm (see Section 5.4 for discussion of allomorphy). This type of analogy that reduces the number of allomorphs is called analogical levelling. Many verbs in English have undergone analogical levelling by changing from a paradigm with multiple root allomorphs to just one allomorph. For example help used to have three root allomorphs in Middle English (help, halp, holp), but these underwent analogical levelling in Modern English so that there is now only one root allomorph (help), by analogy with many other verbs, like float and hope, as shown in Table 14.5.
|Middle English||Modern English|
More rarely, analogy can result in increasing the number of root allomorphs by analogical extension of some other paradigm. For example, the Old English verb hring ‘ring’ had a paradigm with only one root allomorph (hring), but it shifted throughout Middle English and into Modern English to have three root allomorphs (ring, rang, rung), following the paradigm of verbs like begin and sing, as shown in Table 14.6.
|Old English||Modern English|
Analogy can change the inflectional affixes in a paradigm instead of the root morpheme. For example, some nouns in Middle English could form their plural by adding –n. This –n is still retained in a few modern words like oxen and children, but it was replaced in Modern English with –s for nearly all other nouns: eyen > eyes, schon > shoes, treen > trees, etc. This change was due to analogy with the plural pattern with –s that other Middle English nouns followed: cloudes ‘clouds’, foxes, heedes ‘heads’, etc.
The morphological structure of individual words or expressions can also undergo change. One type of change to morphological structure is reanalysis or rebracketing, where an existing morphological boundary shifts across one or more phones. The result is that the word or expression still contains the same morphemes, but some of the phones have shifted from one morpheme to another.
This happened in the history of the English word apron. Middle English naperon was borrowed from Old French naperon ‘tablecloth’, and it continued as napron into Early Middle English. Like many nouns, napron was frequently used with the indefinite article, as a napron. However, the definite article a has a second form an which is used before words that begin with a vowel, as in an apple. By reanalysis, the boundary between the article and the noun was shifted. This caused the [n] to be treated as part of the article rather than the noun, that is, a napron was reanalyzed as an apron. As a result, when the noun was used by itself, it become apron. Interestingly, the reverse boundary shift happened for newt, which was originally ewte, and by reanalysis, an ewte changed to a newte. See Figure 14.8 for a diagram of how reanalysis affected both napron and ewte.
We still see the original [n] in (n)apron in related words like napkin and nape, which did not undergo reanalysis. Similarly, we see the original lack of [n] in (n)ewt in the related word eft, which is used in the term red eft, the name for a juvenile eastern newt (Notophthalmus viridescens).
Another way that the morphological structure of a word or expression can change is if new morphemes or a new structure is imposed upon it that it did not have before. This is called folk etymology or sometimes corruption, in which users of a language treat a word as having an etymology that it does not actually have, and this changes how they interpret its morphological structure. They may even invent completely new morphemes to build the folk etymology, and these morphemes could end up being used to create new words later. The folk etymology normally has some combination of matching phonology and semantics to the original word, though sometimes only the phonology matches.
Folk etymology can be seen in the development of the English word woodchuck, which was a late 1600s borrowing from an Algonquian language. The exact source is not known, but a likely candidate is the word [ʊˈt͡ʃeːk] ᐅᒉᐠ (ochek) ‘fisher (Pekania pennanti)’ from Cree, which is a dialect continuum in the Algonquian family, spoken in Canada and the United States.
Since [ʊˈt͡ʃeːk] contains phones used in Enlgish, it could have just been straightforwardly adapted into English with little change. However, its lack of transparent morphology lead to it being reinterpreted as a compound of wood and chuck ‘throw’, giving a possible etymology (an animal that throws wood around, perhaps) with a similar pronunciation: [ˈwʊdˌt͡ʃʌk] versus [ʊˈt͡ʃeːk]. Real woodchucks do not in fact do anything special with wood, let alone throw it, but the phonological similarity and vague semantics were enough to reinforce this folk etymology. Other examples of folk etymology in English include the following:
- Old English angnægl ‘hangnail’ was originally a compound of ange ‘tight; painful’ and nægl ‘nail’, but it was eventually reanalyzed through folk etymology as hangnail, through influence from both phonological and semantic similarity to hang, since hangnails are pieces of torn flesh that “hang” next to the nail
- Spanish cucaracha ‘cockroach’ was borrowed into English and reanalyzed as cockroach, a folk etymology influenced by phonological similarity to the existing words cock ‘rooster; male chicken’ and roach ‘fish in the genera Hesperoleucus or Rutilus’, which also refer to animals, though very different, so there is only weak semantic similarity; cockroach is now often clipped to just roach
- Old French meisseron ‘mushroom’ was borrowed into Middle English and reanalyzed as mushroom, a folk etymology influenced by phonological similarity to the existing words mush and room, with no notable semantic influence
Folk etymology likely begins at an individual level, often due to a misperception of the real word, when it is sometimes called an eggcorn (Liberman 2003). The term eggcorn itself is also an eggcorn, in this case, a misinterpretation of acorn. Examples of eggcorns include the following, some of which are more commonly used and accepted than others:
- chomping at the bit < champing at the bit
- cold slaw < coleslaw
- for all intensive purposes < for all intents and purposes
- hone in < home in
- wet your appetite < whet your appetite
Many eggcorns are only obvious in spelling, because the replacement is pronounced the same as the original, as in the following:
- deep-seeded < deep-seated
- free reign < free reign
- of < -’ve (as in could of, should of, etc.)
- peak my interest < pique my interest
- tow the line < toe the line
Importantly, folk etymology and eggcorns are fundamentally the same process, just with different levels of acceptance: eggcorns are used by individuals, folk etymology are used by an entire language community. A person is not less intelligent or less competent at English for using cold slaw or could of instead of coleslaw and could’ve. Indeed, they are demonstrating the same creative insight into English that speakers used centuries ago for words like woodchuck and hangnail. As with any case of variation, their version of the language may be mocked now, but some day, it could eventually become the accepted version used by everyone, transitioning from an eggcorn into a folk etymology.
A similar concept is a mondegreen (Wright 1954), which is like an eggcorn in that it is also an idiosyncratic misperception by individual language users. The difference between an eggcorn and a mondegreen is their different effects on the meaning. With an eggcorn, the fundamental meaning is still the same: eggcorn still refers to an acorn, cold slaw still refers to coleslaw, etc. But with a mondegreen, the semantics change, too.
The original example of a mondegreen (and the name of the phenomenon) comes from a 1954 article by Sylvia Wright. In her article, Wright discusses her own childhood misunderstanding of the lyrics of “The Bonny Earl of Murray” (a.k.a. “The Bonnie Earl O’ Moray”), an old Scottish ballad from a collection of songs, poems, and ballads published by Thomas Percy in 1765. The first stanza of the original as published by Percy (1765: 211) is on the left in Table 14.7, with a modernized version on the right:
|Ye highlands, and ye lawlands,
Oh ! whair hae ye been ?
They hae slaine the Earl of Murray,
And hae layd him on the green.
|You highlands, and you lowlands,
Oh, where have you been?
They have slain the Earl of Murray,
And have laid him on the green.
Wright misheard the final line as and Lady Mondegreen, imaging that Lady Mondegreen was perhaps the Earl of Murray’s wife, and both had been killed. This changes not only the pronunciation and morphology, but also the meaning. Wright’s mondegreen includes someone who did not even exist in the original!
Mondegreens are often the result of mishearing lyrics from a song or poem. Perhaps one of the most famous mondegreens is scuse me while I kiss this guy as a mishearing of scuse me while I kiss the sky from the 1967 song “Purple Haze” by the Jimi Hendrix Experience. Jimi Hendrix singing the relevant lyric can be heard starting at 0:45 in the video linked here: https://youtu.be/WGoDaYjdfSg?t=45. Note that the mondegreen changes the meaning of the lyric to something that is more physically plausible (kissing a guy versus kissing the sky), which is why it so easily arises.
Just as an eggcorn may become frequent enough for everyone to adopt it as part of the language as a folk etymology (as happened with woodchuck), a mondegreen may become so common that the standard version of the song changes to match the new misheard lyrics. For example, in the first published version of “The Twelve Days of Christmas” (in a children’s book of unknown authorship from around 1780), the fourth present given to the singer was four colly birds, where colly is an older word for ‘black’ related to coal. However, by the 1909 version arranged by Frederic Austin, colly had become an obsolete word, and the lyric had shifted to the mondegreen four calling birds, no longer referencing their colour at all. Unlike with an eggcorn becoming a folk etymology, there is no special name for a mondegreen like this that becomes standardized.
Folk etymology can also result in back formation, which is a special type of neologism that creates a new morpheme from the false morphological structure, in which the new morpheme is used beyond the original word it was created for. The new morpheme could be a free morpheme or a bound morpheme, depending on the situation.
For example, English borrowed the word emotion from French, but because there are many English nouns having the structure of verb plus the suffix –(t)ion, emotion was incorrectly analyzed as having the same structure: emote–ion. This resulted in the creation of the new free morpheme emote in English. Many other verbs were created in English the same way: donate, resurrect, secrete, etc.
Back formation can also create new bound morphemes, like the English suffix –holic (or –oholic or –aholic), which is back-formed by treating the word alcoholic as if it had the structure alco-holic or alc-oholic, even though it has no such structure (it is actually structured as the Arabic loanword alcohol plus the adjective suffix -ic). This invented morpheme -holic is treated as meaning ‘addict’, and it can be used fairly productively to create new words, like chocoholic ‘chocolate addict’, shopaholic ‘shopping addict’, sugarholic, ‘sugar addict’, workaholic ‘work addict’, etc.
A word that uses a back-formed bound morpheme like –holic looks a lot like a blend; that is, we might think of shopaholic as being a blend of shop and alcoholic, rather than being shop plus the back-formed suffix –holic. However, blending usually results in only a single neologism, while a morpheme created through back formation could be used to create many new words. Language is flexible though, and it could be difficult in some cases to draw a clear boundary between blending and back formation.
Other examples of back formation in English include the following:
- some verbs like babysit and edit were back-formed from nouns (babysitter and editor) which are similar to other nouns that are built from a verb plus the suffix –er or –or, like baker and sailor
- some singular nouns like bicep and pea were back-formed from nouns (biceps and pease) which were originally already singular, but they seemed like regular plurals ending in –s, so they were misinterpreted as plurals, requiring new singulars to be created
- the prefix cyber– was back-formed from the word cybernetics and can now be used quite productively to derive new words having to do online activities, like cyberbullying, cybercafe, cyberspace, etc.
Morphology and syntax are intimately related, especially in the realm of grammatical functions, like verb tense and noun case (see Section 5.7 for many others). Some languages may express a particular grammatical function synthetically (as a bound morpheme within the same word as the root), while others may express it analytically (that is, as a free morpheme separate from the root). Some languages can even differ within themselves, expressing some grammatical functions synthetically and some analytically.
For example, in English, the past tense is ordinarily formed synthetically with a suffix, while the future tense is formed analytically with a separate word. Contrast this with Malagasy (a Greater Barito language of the Austronesian family, spoken in Madagascar), in which both the past and future tenses are synthetic (with the tense morphemes being prefixes), and with Māori (a Tahitic language of the Austronesian family, spoken in New Zealand), in which both tenses are analytic (with the tense words preceding the verb). The different patterns for these three languages are demonstrated in Table 14.8, with data adapted from Richardson 1885 (for Malagasy) and from Williams 1917 (for Māori).
|language||past tense||future tense||type|
|English||[pled]||[wɪl ple]||synthetic / analytic|
|Māori||[i taːkaro]||[ka taːkaro]||analytic|
Over time, languages can change from synthetic to analytic and vice versa, because of the connection between morphology and syntax. Morphology and syntax essentially behave as two aspects of the same underlying concept, called morphosyntax. Diachronic changes to the morphosyntax of a language, as well as changes to just the syntax, are discussed in Section 14.5.
Check your understanding
Anonymous. circa 1780. Mirth without mischief. London: J. Davenport and C. Sheppard.
Austin, Frederic (arr.) 1909. The twelve days of Christmas (traditional song). London: Novello & Co.
Klafehn, Terry. 2003. Emergent properties of Japanese verbal inflection. PhD dissertation, University of Hawai‘i.
Percy, Thomas (ed.) 1765. Reliques of ancient English poetry: Consisting of old heroic ballads, songs, and other pieces of our earlier poets, (chiefly of the lyric kind.) Together with some few of later date. Volume 2. London: J. Dodsley.
Richardson, James (ed.) 1885. A new Malagasy-English dictionary. Antananarivo: London Missionary Society.
Shibata, Takeshi. 1965. 生きている方言 [Living dialects]. Green Belt Series 63. Tokyo: Chikuma Shobō.
Williams, Herbert W. 1917. A dictionary of the Maori language. Wellington: Marcus F. Marks.
Wright, Sylvia. 1954. The death of Lady Mondegreen. Harper’s Magazine : 48–51.