The Jungang Journal of English Language & Literature, 45(4), 2003

Semantic Change1) 1)

D. Gary Miller
(University of Florida)

Ⅰ. Introduction : Expressivity and Change

Semantics is the part of the grammar that involves meaning.

Semantic change, like all other change, begins with expressive usage in
a given context, i.e. assignment of new meanings to old words. In 1819,
cram began to be used as expressive university slang for intensive
study, the literal image being one of forcing knowledge into oneself by
a stuffing process. The creative use of hairy to convey contorted
difficulty dates to 1848. Since 1963, jock has been an expressive term
for a versatile young athlete. And so on.
Emphatic language is one typical variety of expressivity. It is
responsible for double superlatives of the type bestest [1868], mostest
[1885]. Note also adjectival intensifiers, such as plumb (crazy) [1587],
dead (right) [1589], filthy (rich) [1616], dreadfully (bad) [1697], awful
(good) [1818], real (sturdy) [1827], totally (awesome) [1972], etc.
(ODS335ff.). As noted by Stern (1931: 394), loss of expressivity is
evident when an intensifier accompanies a word of contrary meaning, as
mighty little, pretty dirty. Loss of expressivity may be regarded as a
subclass of the more general semantic bleaching, variously defined (and
disputed) in the literature. I will take it to designate the reduction or
loss of a distinctive property. For instance, in expressions like the larger

1) For comments on an earlier draft of this paper, I am grateful to Galia Hatav.

- 135 -
136 D. Gary Miller

half or it broke into three halves, a precise mathematical bipartization

into identical portions loses the identity and/or bipartite requirement (cf.
Stern 1931: 406). Unique has shifted from non-quantifiable to
quantifiable in expressions like very unique.
Expressivity lead to many fanciful coinages, such as discombobberate
[1838], discombobulate [1916]; descriptions of speed: lickety-split [a1845];
words for ‘large’ or ‘great’: splendiferous [a1845], bodacious [a1845], fan-
tabulous [a1960], humongous [a1970]; words for nothingness: doodly-
squat [1934], diddly-squat [1963], zilch [a1970], zip(po); terms for an
unnamed or forgotten lexical item: thingamabob [a1770], thingamajig
[a1825], contraption [a1825], doohick(e)y [1914], gadget [a1855], widget
[a1930], gizmo [a1945]. Finally, any malfunction can be described as a
glitch [1962].
Creative exaggeration/hyperbole is responsible for countless descriptive
terms, e.g. for very hot weather: roasting [1768], scorcher [1874], sizzler
[1901], etc., and their opposites (freezing, etc.) for very cold weather.
With loss of expressivity, a once expressive variant can become the
standard. For instance, moonlight [1957] ‘hold an additional evening job’
has become the usual verb for holding more than one job. With the loss
of expressivity, there is a concomitant change in the direction of
prototypicality, e.g. L caballus ‘pack-horse, nag’ in Vulgar Latin was
used expressively for ‘horse’, and in Romance became the standard word
for ‘horse’ (F cheval, Sp. caballo, etc.). L bucca ‘(puffed out) cheek’
replaced ōs ‘mouth’ (cf. F bouche ‘mouth’, etc.); L pellis ‘hide; leather’
replaced cutis ‘skin’ (cf. F peau ‘skin; hide’); and so on (Meillet 1965
[1905-06]: 264). Diminutives deprived of expressivity can replace the base
noun, e.g. ML sōliculus ‘little sun (sōl)’ > F soleil ‘sun’ (Fruyt 1989: 129).

Ⅱ. Extralinguistic Input

Sociocultural history is the part of a word’s etymology that

 Semantic Change 137

determines the semantic changes a word can undergo. German Kopf and
E cup can be related formally, but Kopf means ‘head’ and cup is a
drinking vessel. The link is provided by medieval battles, in which the
smashing of heads was compared to the smashing of (clay/ceramic)
cups. Since OE cuppe derives from LL cuppa ‘drinking vessel’, Germ.
Kopf as ‘head’ was originally a metaphor based on the approximate
shape and smashability (cf. Bloomfield 1933: 440; Fritz 1998a: 108f., both
w. lit). Similar is F tête ‘head’ (§3.2).
Cultural change can be illustrated by the different uses of dinner and
supper over time (Stern 1931: 196f., w. lit). Dinner was originally eaten
around 9 A.M. It was etymologically ‘breakfast’: VL *disjējūnāre (cf.
ML dējējūnāre) ‘break the fast’ (L jējūnium ‘fast(ing)’) > OF disner
b
[c12 ] (cf. ML disnāre) > F dîner DINE. Dinner gradually got later,
closer to noon by c17 (cf. F déjeuner ‘lunch’; the older meaning
‘breakfast’ is now generally subsumed under petit déjeuner ‘little lunch’).
Supper (OF soper [1175]) was the late-afternoon meal. By 1782 dinner
was eaten around 3 P.M., and then became the evening meal, relegating
supper to a light meal after the theater. Regional variation in these
terms continues today.
Another culturally-induced change involved the addition in early
Roman times of the months January and February at the beginning of
the year, which originally began with Mārtius ‘of Mars’ (> March),
Aprīlis (? Etruscan?), Maius ‘of (the goddess) Maia’, Jūnius ‘of Juno’
(cf. Varro, De Lingua Latina 6.33; Ovid, Fastae 1.39; etc.). This entailed
the shift of Quīn(c)tīlis, Sextīlis, September, Octōber, November,
December, from the fifth through tenth months, respectively (cf. quīnque
‘five’, sex ‘six’, septem ‘seven’, octō ‘eight’, novem ‘nine’, decem ‘ten’), to
the seventh through twelfth months. Quīn(c)tīlis was renamed Jūlius
JULY in -44 to honor Julius Caesar, and Sextīlis was changed to
Augustus AUGUST by the Emperor himself in -8. Just why the first
two months (Jānuārius ‘of Janus’, Februārius ‘of februa ‘festival of
138 D. Gary Miller

purification’) were added to the Roman calendar is a complete mystery.

It is also unclear when this innovation occurred. Based on the fact that
the Roman consuls entered office on January 1 for the first time in
-153, it is speculated that that was the year the two additional months
were added, but that is not conclusive and many unknowns remain
(discussion in Gordon 1983: 228f.).
Political correctness, a special kind of psychosocial force, drove the
replacement of homosexual by gay. Initially, gay acquired the meaning
‘homosexual’ as an expressive variant with positive connotations around
1915. Given the negative and condemnatory connotations of homosexual,
the non-judgmental term gay was preferred in some circles.
Subsequently, political correctness elevated its status to standard usage.
Cognition plays a role in most semantic change, but one specific
aspect involves mental links between actors, instruments, activities, and
locations (discussion in Miller 1993: 68ff.; Bauer 2002). A planter, for
instance, can be a person who plants (actor), an object that plants
(instrument), or a place for plants (location). Many suffixes exhibit this
polysemy (cf. Serbat 1975: 360-75). Since activities presuppose places,
action nouns frequently evolve into location nouns. For example, L
pr(eh)ēnsiō ‘act of apprehending’ came to mean ‘captivity’ and then
‘place of captivity’, as in F prison PRISON (Nyrop 1913: 223). Words
that remain ambiguous between the activity and location senses are
study, passage (ibid., p. 216). Since agent, instrument, and means
constitute the intermediary through which a process can be realized,
activities can evolve into agents or instruments. Trap and press, which
began as activities, today are instruments.
 Semantic Change 139

Ⅲ. Specialization and Generalization

1. Polysemic Extension and Limitation

A word will typically acquire an expressive meaning in a particular

subculture, whence it may re-emerge with this new meaning in the
speech community at large. Fritz (1998a: 31; 1998b: 870) calls this
Meillet’s Generalization. For instance, arrive (< French) goes back to
Late Latin *ad+rīp-ā-re ‘to reach the shore’ (rīpa ‘bank, shore’). In
sailor’s jargon the meaning was extended to mean ‘reach (any place)’
because when sailors reached shore, that was, broadly speaking, their
prototypical destination (Meillet 1965 [0905-6]: 259; see also Nyrop 1913:
249ff.). English borrowed arrive from French with the generic meaning
(cf. Görlach 2000: 106).
Many Greek examples are documented by Hecht (1888: 52-61), in
which a word develops a specialized meaning in a subculture, and that
meaning emerges as (at least one) standard usage. For instance,
sullogismós originally meant ‘computation; reasoning’, but in the
language of philosophy developed the specialized meaning SYLLOGISM.
G krãsis ‘mixing, blending’ in the language of the grammarians came to
designate a special kin of contraction (CRASIS). G ónux ‘hoof; claw;
nail’ in earth language meant ONYX (maybe as early as Mycenaean:
Hooker 1980: §188). G krústallos in Homer meant ‘ice’; in the language
of earth science, it came to mean (rock-)CRYSTAL’.
Salary goes back to F salarie, from L salārium ‘(money) to buy salt
(sāl)’. Roman soldiers were paid a salārium to buy salt, which was
expensive. In Anglo-French, the word for payment to a soldier or an
employee was salarie, which came into use in Late Middle English, the
restriction to salt (the original meaning!) having been lost. For the value
once attributed to salt, compare the expression (not) worth one’s salt.
Extension/widening was once formalized as loss of a semantic feature.
Bird originally meant a young or small fowl, and lost the feature
140 D. Gary Miller

[+young] (Williams 1986: 175f.). Holidays are no longer ‘holy days’, but
include Labor Day, days off from school, etc., and have apparently lost
the feature [religious], optionally at least (ibid. p. 175). More complicated
changes, which cannot be described/formalized as simple loss of a
feature also occur. Apéritif began as a medical term for a substance to
open the pores and got generalized to a drink that opens the appetite
(Nyrop 1913 §127). In an analysis that accounts for both types, the
range of application of the new meaning is a subset of the range of the
old meaning (Geeraerts 1997: 895; cf. Ullmann 1957: 117).
Narrowing/restriction was formerly analyzed as addition of a semantic
feature or specification. Liquor [-solid, +fluid] → [-solid, +fluid,
+alcoholic] (Williams 1986: 171). E fowl originally meant ‘bird’ (like
Germ. Vogel), but became a specific kind of bird (Ullmann 1962: 229). L
doctrīna ‘instruction’ > DOCTRINE ‘lesson; teaching’ [c14-c17e], then
came to be restricted to ‘religious dogma’ (Copley 1961: 59), formalizable
by addition of a feature [+religious]. OE dēor designated a ‘fourfooted
animal as object of the chase’ (excluded domestic animals and birds).
Probably among hunters, dēor came to be specialized as Cervidae, the
animal of the chase par excellence. The specialization occurred in Old
English, but the more general meaning persisted into Middle English,
where it was gradually supplanted by the loanwords beast and then
animal, rendering the more general meaning of DEER obsolete (Stern
1931: 416). A formalization like dēor [+quadruped, +hunted] → deer
[+quadruped, +hunted, +Cervidae] is not very insightful because, at every
stage, “the new range includes the old one” (Geeraerts 1997: 95).
One of the problems with formalization by loss or acquisition of a
semantic feature is that many examples can be formalized as a change
from an underspecified feature to a specified feature. For instance,
corpse derives from F corps ‘body’, which can be alive or dead, i.e.
underspecified for the feature [living] (viz. [uliving]). The restriction to
[-living] can be formalized as [uliving] → [-living].
 Semantic Change 141

Another problem involves the conception of the feature-composition of

a given word. What are the relevant features for desert (L dēsertum
‘deserted; abandoned’)? There was supposedly a change in c19 from
‘uninhabited/uncultivated region’ to ‘barren, sandy waste’ (Copley 1961:
51f.). Now, anyone who has seen the deserts in the southwestern United
States knows that ‘barren, sandy waste’ does not come close to an
accurate description. Yet, most of us have a prototypical schema
featuring an Arabian kind of desert, which can be so described. Was
there a restriction in meaning for the word desert, or just for our desert
schema? And if we cannot agree today on what a desert is, how can
we be sure of how deserts were conceptualized in the past? Since
Middle English, desert has designated a wasteland, but also in some
texts included forests, retreats, and some other things. Apart from that,
even the idea of a wasteland is not clear. Maybe the word has not
changed meaning. Perhaps it has always designated at the core
whatever one’s conception of a barren wasteland might look like. And
what is the feature composition for that?
Borrowing is a frequent source of semantic restriction. The range of
red and yellow decreased with the borrowing of orange (Görlach 2000:
102). A similar case from Basque is reported by Trask (1996: 46):
Basque urdin originally covered the range of blue, green, and grey, but
came to be restricted to ‘blue’ with the borrowing of berde ‘green’ and
gris ‘grey’. More examples are discussed in Fritz (1998a: 139f).

2. Semantic Shifts

Specialization (or restriction) and generalization (or extension) can

work together and produce the effect of a shift. Consider (1), from
Bloomfield (1933: 430ff.).
142 D. Gary Miller

‘edible part of ‘muscular part

STAGES ‘nourishment’ ‘edible thing’
animal’ of animal’
Ⅰ food meat flesh flesh
Ⅱ food meat meat flesh
Ⅲ food food meat flesh
Ⅳ nourishment food meat flesh
Stage Ⅰ occurs in Old English. In Middle English (Ⅱ) the meaning of
meat was generalized to partly overlap with that of flesh whose
meaning was then restricted. Later, food encroached on meat, and
replaced it in its original meaning. Still later, nourishment, an
Anglo-French construct, encroached on food and ousted it from its
original sense.
Consider the fate of L testa ‘pot’, caput ‘head’ and crānium ‘skull’ in
French (Benveniste 1966 [1954]: 295f.). Crānium survives as F crâne in
the same meaning. Caput in Latin had the literal and many figurative
meanings of ‘head’ (Moussy 1989), but survives as F chef only with
figurative meanings. The literal meaning was taken over by testa ‘pot’
→ ‘pot; skull’ → ‘skull; head’ → F tête ‘head’ (cf. Fritz 1998a: 109).

Ⅳ. Melioration and Pejoration

As a special kind of restriction, a word may acquire a positive value

(melioration) or a negative connotation (pejoration). Accident was
originally neutral (‘any happening’), but became negative, as did casualty
and fate. By contrast, the originally neutral fortune, luck, and success
acquired a positive valuation (cf. Nyrop 1913: 144-8; Tournier 1985:
290f.). Drug tends to have negative connotations unless qualified by
prescription. Cunning originally meant ‘knowledgeable; skillful’ and was
a positive term. Around the end of c16, it became more negative
 Semantic Change 143

(‘crafty; sly; devious’), and today is mostly pejorative (Copley 1961: 47;
Ullmann 1962: 174). Other examples follow.

Degeneration/Pejoration
a) despot < G des-pót-ēs ‘head (lord) of the household’ who had
absolute power and control over his family.
b) villain < ME/OF vilain ‘feudal serf’ < ML vīllānus ‘one who
works at a villa/farm’ became notorious in other ways, hence
‘outlaw’ (Nyrop 1913 §180).
c) cheat(er) < escheat(er) ‘property reverting to the state by escheat
(< L ex-cadere ‘fall out’)
d) OE hūswīf ‘woman of the house’> hussy
e) OE cwene ‘woman’> quean ‘whore’

Elevation/Melioration (cf. Ullmann 1962: 234f.)

a) OE cniht ‘boy, knave, servant’ > knight
b) OE hlāf ‘loaf’ + weard ‘keeper’ > hlāford > lord
c) OE stiʒe-weard ‘sty-keeper’ > steward
d) L minister (< minus ‘less’) ‘attendant; servant’ was elevated first
in Christian contexts, then especially when ministers became
counselors and representatives of the prince (Nyrop 1913 §§81,
192f.).
e) LL cancellārius ‘usher stationed ad cancellōs (at the grating /
CHANCEL) of a basilica or other law court’ > chancellor. The
change occurred in the Eastern Empire, where this officer became
a secretary or notary, and later a court official. In England, under
the Norman kings, the status of this official was elevated even
more (Ullmann 1962: 234, w. lit).

The etymology of boy is disputed, but at least some elevation/

melioration is involved; cf. the proposed lexical entry by Liberman (2000:
144 D. Gary Miller

224):

BOY, 1260. Original meanings ‘churl, servant’, ‘devil’ (rare); regularly

used as a derogatory word. The meaning ‘male child’ does not occur
before 1400. Apparently, a blend of an onomatopoeic word for a noisy
evil spirit (*boi) and a baby word for ‘little brother’ (*bo). The latter
can be extant in the proper name Boia (OE). Both words have
numerous counterparts in and outside English and Germanic.

Consider the changes in churl/carl (Rinelli 2001). OE ceorl ‘peasant;

(free)man’ (by contrast to eorl ‘nobleman’ and þēow ‘slave’) in Early
Middle English was demoted to ‘serf; bondman’ [1225+]. Around 1300,
the technical reference to the social class became an evaluative term
‘one of rude manners; rustic, boor’, then ‘base fellow; villain’. The first
change was prompted by the blurring of the lower classes, due to social
advancement of the slaves: “ME cherl came to be applied also to former
slaves, placed now on the same level as churls” (Rinelli, p. 269).
Because the former slaves “still retained rude manners” (ibid.), there
was a further decline of churl. In northern texts, Nordic carl (ON karl
‘man; commoner’) was generally a term of address (‘fellow’) because
peasants there had higher status, while churl was a mostly
contemptuous term for ‘slave; thrall’. In the London area, the opposite
valuation occurred, and carl is the more contemptuous term.
The changes in this section are based on the tendency for objective
labels to become subjective and/or evaluative (cf. Traugott 1989: 34f.).
Traugott treats the subjectivization of modals from deontic (denoting
obligation, permission) to epistemic, involving the speaker’s knowledge or
belief. For instance, she cites the change from you must do that, where
deontic must signals an obligation, to politicians must be living on the
moon, where epistemic must expresses a subjective judgment (cf.
Sweetser 1990: ch. 3; Kövecses 2002: 216ff.). Fritz (1998a: 55f.) relates
 Semantic Change 145

the change from wishes to commands.

Ⅴ. Metaphor

In its simplest form, the essence of metaphor is that X resembles Y.

For instance, labyrinth is frequently applied to any complicated building
with an intricate network of hallways. More formally, “Metaphor is the
cognitive mechanism whereby one experiential domain is partially
‘mapped’, i.e. projected, onto a different experiential domain, so that the
second domain is partially understood in terms of the first one”
(Barcelona 2000b: 3). More simply, “one concept (the target) is
structured (understood) in terms of the other (the source)” (Feyaerts
2000: 60). The mapping of one domain onto another can be illustrated
with the conceptual metaphor THEORIES ARE BUILDINGS: they are
constructed by laying foundations and building on other theories, but
ultimately get demolished (Kövecses 2002: 108, 130f.).
Hyperbolic metaphors, such as splitting hairs for detailed subdivisions,
paradise for any pleasant place, and iceberg for an unemotional person,
are very frequent (Stern 1931: 310ff.). The last is opposite of the
EMOTION IS HEAT (OF FIRE) metaphor (Kövecses 2002: 113): fiery
people have hot tempers and burn with emotion. It is interesting that
both love and anger share the fire metaphor (Kövecses, pp. 203ff.): an
angry person is fuming, burned up, smoking from the ears, red in the
face, and erupts like a volcano; ardent lovers carry a torch and burn
with passion until the flames die down and the fire is out.
Sensory (synaesthetic) metaphors include the following: piercing
(sounds), loud (colors), sweet (personality, voices, perfume), warm
(colors, scent), hot (music, mustard, people), hard (liquor, rock), soft
(drink), etc. (cf. Ullmann 1957: 233, 266-89; 1962: 216; Tournier 1985:
243ff.). Sweetser (1990: 29) explains the metaphor of bitter anger as
146 D. Gary Miller

follows: “the anger is unpleasant to our emotions in a way analogous to

that in which a bitter taste displeases our tastebuds.” She goes on (p.
30) to note that “bodily experience is a source of vocabulary for our
psychological states, but not the other way around.”
Body parts are particularly prone to metaphorical extension (Nyrop
1913: 229ff.; Guiraud 1980, chap. 3). Some obvious metaphors are head
of cabbage, column, committee, etc., mouth of a river, face of a watch,
eye of a needle, nose of an airplane, teeth of a saw, hand of destiny,
arm of justice, foot of a mountain (Piedmont = F pied du mont ‘foot of
the mountain’), hearts of celery, arms, legs, back of a chair, back of a
book, neck of a bottle (bottleneck), electronic brain. OF/ME dent-de-lion
‘lion’s tooth’ > DANDELION (from its sharply indented leaves). THE
STRUCTURE OF AN ABSTRACT COMPLEX SYSTEM IS THE
PHYSICAL STRUCTURE OF THE HUMAN BODY (Kövecses 2002:
129f.): the head of state may not be able to cure an ailing economy,
which is symptomatic of deeper illnesses that can penetrate to the heart
of society and destroy its very backbone, rendering it impossible for any
governing body to put it back on its feet.
Social forces share the same metaphors as mechanical forces:
attraction, impulse, inertia, momentum, propulsion, resistance, etc. (Stern
1931: 294).
Personal attributes and insults are frequently animal metaphors, e.g.
ass, bitch, chameleon, chicken, crab, dodo, dog, fox, gnat, hawk, hog,
jellyfish, nag, ox, pig, rat, shark, skunk, snail, snake, stag, swine, tiger,
turkey, turtle, vermin, vixen, vulture, wolf, worm. The traits associated
with some 140 animals are listed in Tournier (1985: 238-43); cf. de la
cruz Cabanillas and Tejedor Martinez (2002: 244-50). Note also animal
and beast themselves as applied to humans. Animal metaphors can be
qualified: (jack)ass, (old) bat, (fat) cat, (fat) cow, (real) dog, (silly)
goose, (big) horse, snake (in the grass), (bunch of) sheep. They take
adjectival suffixes (bearish, catty, foxy, hawkish, loony, sheepish,
 Semantic Change 147

squirrely), and convert to verbs: ape, bitch, bug, dog, hog, hound, nag,
parrot. Some animal verbs acquire a particle: ferret out [1577], clam up
[1920], horse around [1928], chicken out [1934], rat (out) [1935], weasel
out [1962], pig out [1978]. Most of these instantiate the conceptual
metaphors OBJECTIONABLE BEHAVIOR IS ANIMAL BEHAVIOR and
OBJECTIONABLE PEOPLE ARE ANIMALS (Kövecses 2002: 125ff.).
Parts of animal bodies are frequently transferred disparagingly to
human parts: beak, bill, paw, proboscis, etc. Small animals, such as
lamb, kitten, can be terms of endearment.
Animal names can be equally applied to instruments. An old metaphor
involves the crane, with its long neck, applied to a lifting machine. G
géranos ‘crane’ was also used thus (Hecht 1888: 52). Its diminutive
geránion [‘little crane’] was applied to a plant whose fruit resembles a
crane’s bill. The Romans borrowed the word as geranium GERANIUM.
The bone at the base of the spine resembles the beak of a cuckoo and
bears its name COCCYX (G kókkux ‘cuckoo; coccyx’) (Skoda 1988:
224f.). From G polúpous ‘many footed; octopus’ is derived POLYP
(Skoda, P. 232).
Metaphors are a frequent source of compounds for new concepts:
Milky Way, railroad, skyscraper (Ullmann 1957: 183). Many compound
metaphors are (neo) classical, e.g. astronaut (‘star-sailor’), psychedelic
[1956] (‘mind-clear’).

Ⅵ. Metaphorical Change

To figure out a given change, the comparison may require a little

imagination and even extralinguistic (real world) knowledge, but remains
essentially a comparison.
Examples:
a) window < ON vind-auga ‘wind-eye’
148 D. Gary Miller

b) butter < G boú-tūron ‘cow-cheese’ (cf. G tūrós ‘cheese’)

c) galaxy [Ch.] (< G galaxíās (kúklos) ‘milky (circle)’) is slightly
more complicated in that the comparison to milk is a metaphor, but the
transfer of ‘circle’ involves metonymy (see below).
d) dactyl (< G dáktulos ‘finger’) as a metrical foot type has a long
constituent followed by two short, analogous to the joints of a finger.
e) electricity (< G ḗlektron ‘amber’) requires the extralinguistic
information that amber produces sparks when rubbed together, but it
remains a simple comparison (metaphor). To figure out the link one
need only know what amber is/does, and what electricity is/does.
Color words frequently evolve from items which are prototypically
that color (Fritz 1998a: 140), e.g. violet (L viola ‘violet plant; violet
color’), rose (L rosa ‘a rose’), orange (< Arab. nāranj < Pers. nārang <
Skt. nāranga- ‘orange (tree)’), aubergine (< Catalan alberginia < Arab.
al-bāḏinjān < Skt. vatin-ganah-).
Unrelated words can coalesce through phonological change or
historical accident, such as borrowing (Bloomfield 1933: 436; cf. Ullmann
1957: 128ff.; Meier 1989: 64f.). Ear of corn is etymologically not the
same word as the body part ear, but fits into the general pattern of the
metaphorical use of body parts, this time by folk etymology. This type
of metaphorical reinterpretation due to homophony is sometimes called
FUSION. Other examples include corn ‘horny hardening of skin’ and
corn ‘grain’, or fuse ‘electric circuit protector’ (Ital. fuso < L fūsus
‘spindle’) and fuse ‘detonator of an explosive’ (L fūsus, PPP to fundere
‘to pour; melt’). In these cases, unrelated words merged into one. It can
also happen, of course, that related words diverge. Only etymologists
would relate sole ‘bottom of a shoe’ to the name of the fish, the
adjective long to the verb to long (Ullmann 1962: 178), or staff ‘stick;
rod; wand’ to staff ‘workers assisting a manager’.
 Semantic Change 149

Ⅶ. Concrete > Abstract

Based on the idea that abstract concepts arise cognitively from bodily
experiences (Goddard 1998: 78ff.), abstract meanings evolve from
concrete by metaphor; cf. field (of math, etc.) < cornfield, etc.; scale
(e.g. musical) < L scāla(e) ‘stairs; ladder’, by the similarity of shape;
climax < G klîmax ‘ladder’, by reference to the function of a ladder in
helping one “reach the top.” Eliminate in Latin had only its concrete
meaning: put (someone) ‘out of’ (ē/ex) the ‘house’ (līmen ‘threshold’); cf.
Bloomfield (1933: 430).
Understanding is frequently conceptualized as a metaphor of grasping
or seizing (cf. Bloomfield 1933: 429f.; Ullmann 1962: 225; Sweetser 1990:
28): L comprehendō ‘press together; get hold of; COMPREHEND’; cf.
grasp [1680], catch on [1884], get [1907] ― all in addition to earlier
literal meanings. In other words, physical grasping gives way to
cognitive grasping (Williams 1986: 181; Sweetser 1990: 38). The
metaphor underlying understand involves standing under; cf. the
expression get to the bottom of (something).
At least in the sensory domain, concrete > abstract (perceptive/
cognitive) > metalinguistic (expressive) > subjective, as in the evolution
of wit (Koivisto-Alanko 2000a, b). In 1350, wit meant ‘five senses’
(perception) and ‘mind; seat of consciousness’ (cognition). By Early
Modern English, the perception meaning disappeared, replaced by sense.
Also, wit acquired expressive new meanings, such as ‘superior
intelligence’. In Tillotson [a1694], wit has the modern expressive sense.
In Burnet [a1715], it is personified (subjectified) by metonymy (see
below).
The change from perception to cognition (cf. §4) is also exemplified
by the PHYSICAL VISION ⇒ MENTAL “VISION” metaphor (Sweetser
1990: 33), as in “I see” = ‘I know’, and several examples adduced by
György László (apud Kövecses 2002: 219): fantasy (< G phantasíā
150 D. Gary Miller

‘appearing; appearance; imagination; representation’); idea (< L idea

‘archetype; idea’ < G idéā ‘form; semblance; archetype; idea’); intuition
(< L intuērī ‘to look at; observe; contemplate’); etc.: “vision is connected
with intellection because it is our primary source of objective data about
the world” (Sweetser 1990: 39).
Spatial notions are generalized to temporal by metaphor. For instance,
second (unit of time) derives from the idea of following sequentially in
space.
One of the ways in which concrete terms can evolve into abstract is
by construal with abstractions. For instance, very concrete affective
verbs can be used figuratively of abstractions (Nyrop 1913: 231f.):
wound the soul, break the silence, underscore a question. From such
contexts, abstract figurative meanings can be generalized, entailing that
a wounding need not involve physical assault, an underscoring need not
involve an actual line, and so on.
The evolution is complete when the original concrete meaning is lost
entirely. For instance, L errāre originally meant ‘to wander, stray, roam’
in the concrete sense, but already in Early Latin could mean ‘be
mistaken’. Both meanings were continued in OF errer and E ERR from
Chaucer to the end of c17, when the literal meaning was lost (Copley
1961: 64f.). English insult goes back to L īnsultāre ‘to leap upon; attack’
(Copley p. 92); offend is from L offendere ‘to knock against; strike’
(Copley, p. 112). These and many similar examples (debate, contend,
convince, etc.) illustrate the SPEECH IS COMBAT metaphor (Sweetser
1987: 454).
It is of course not the case that abstract nouns derive only from
concrete by metaphor. It is well known (cf. Nyrop 1913 §328) that there
is no language in which abstract nouns cannot be derived directly by
means of some morphological operation.
Finally, the change is not unidirectional. Changes from abstract to
concrete are also attested. OE ʒeogoð (YOUTH) was derivationally an
 Semantic Change 151

abstract noun ‘youngness’ with the IE suffix *-ti- > Gmc. *-þi- (cf.
Krahe-Meid 1967: 156; see *yeu- in Watkins 2000: 103), but already in
Beowulf the concrete collective meaning ‘young persons; warriors’ is
also found (cf. Stern 1931: 351f.). This involves metonymy (below).
Other examples of abstract > concrete include beauty, failure, help,
horror, curiosity, celebrity, etc. (cf. Nyrop 1913: 224f.; Stern 1931:
370ff.). Abstracts in -ing frequently evolve into concrete nouns:
crossing, writing, painting, covering, dripping(s), clipping(s), cutting(s),
holding(s), earings(s), shavings(s), etc. (Miller 2002: 316, w. lit).

Ⅷ. Metonymy

Two major approaches to metonymy share the idea “A stands for B”

(Feyaerts 2000: 62). More specifically, X is associated to a conceptually
proximous Y. This is the traditional approach, defended by Feyaerts
(2000). It is reasonable to interpret metonymy as SEMANTIC
REANALYSIS, reinterpretation of a word with meaning X as meaning
Y (or X + Y) from an associated context. The association is by
conceptual contiguity (cf. Bain 1893: 186-95) and requires extralinguistic
information to make the necessary link(s). All metonymy involves
transfer of meaning in some way, either from one word to another
closely associated word, or, more generally, by means of an associated
social/cultural context. Following is the classic definition of metonymy,
as stated by Stern (1931: 298):

Metonymy ... is founded on contiguity, and consists of naming a thing

by some accompaniment. The accompaniment may be (1) the Sign or
Symbol, or any significant adjunct, (2) the Instrument for the Agent,
(3) the Container for the thing contained, (4) an Effect for a Cause,
(5) a Maker for his Works, (6) the name of a Passion for the name
152 D. Gary Miller

of its Object. Closely related to these forms of metonymy are the

forms of synecdoche founded on contiguity: (1) Naming a thing by
some Part, (2) the reverse operation of using the Whole for the Part.

Several types of metonymy are commonly recognized in the literature.2)

Ⅸ. Transferred Epithet

An epithet (G epítheton ‘added; assumed’) is a name or label acquired

by virtue of some trait or accomplishment. It is an associated modifier,
as in Richard the Lionhearted. As a form of metonymy, the meaning of
the head is transferred to an associated contiguous modifier. Stern (1931:
248) illustrates with a standard elliptical expression “sweet or dry?.” If
this were generalized to non-restaurant contexts, so that in standard
English, sweet meant ‘sweet wine’, that would be an example of
transferred epithet; cf. private = private soldier, shrapnel = Shrapnel
shell, both generalized from a military context (Stern 1931: 274f.).3)
Examples of metonymic change by transferred epithet include:

2) Not all scholars define metonymy the same way. For some, the definition is
narrower; others restrict it to synecdoche. For our purposes, it does not
matter whether one includes all of what follows under the heading metonymy
or subdivides it into separate categories.
3) Given such examples as ten-speed [bicycle], the question has rightly been
asked whether these indeed involve a transferred epithet or simple clipping.
Clips such as penult, decaf invariably preserve the meaning of all components.
If there is any distinction, it might be that clipping destroys some part of the
remaining constituent. By that criterion, one of the classic examples of
metonymy, roasted meat → roast, would actually be an example of clipping
because the -ed of roasted is also deleted. Having asserted that it is not clear
that there is any substantive difference, I will nevertheless follow the criterion
that in standard metonymy by transferred epithet no part of the remaining
component is clipped. (Fritz (1998a: 51) subsumes all examples like F voiture
e
automobile [c19 ] ‘self-moving vehicle’ → automobile ‘id.’ under “Elliptical
Uses” (cf. Geeraerts 1997: 101).
 Semantic Change 153

a) Flour began as the same word as flower, specified as flower of

wheat, flower of barley, etc. In 1691, there is a reference to Milk, Water,
and Flower, seasoned with Salt. Already Genesis and Exodus 1013
[c.1250] reads: kalues fleis, and flures bred, and buttere ‘calf’s flesh
(meat) and bread of flour and butter’ (cited by Stern 1931: 272).
b) G galaxíās (kúklos) ‘milky (circle)’ (→ GALAXY by transfer of
‘circle’ to the ‘milk’ word.
c) L sermō religiōsus ‘religious speech’ > LL sermōn(e)- (religiōso-)
(→ SERMON, in which the word for ‘speech’ also bears the sense of
‘religious’.
d) L nātālem [diem (Jēsūs Christī)] ‘birth [day (of jesus Christ)]’ > F
Noël (NOEL).
e) L via strāta ‘roadway spread over (with stones or the like)’ → LL
strāta [c4] ‘paved road’ (> STREET)
f) L (*movimentum >) mōmentum temporis ‘movement of time’, then
‘(brief) space of time’, and the notion of time was transferred to
mōmentum (> MOMENT).
g) L Cyprium(aes) ‘(copper) of Cyprus’ → LL cuprum [= cu on the
Periodic Chart] (> copper), in which the original word for Cyprus bears
the meaning of its export. Copper was associated with Cyprus, and it is
that (non-material) association that makes the change an example of
metonymy. It is not a metaphor because copper does not resemble
Cyprus!
There are many examples of toponyms imparting their name to an
associated product (Nyrop 1913: 391-5). Edible products include fruits
(cantaloup), cheeses (brie, camembert, gruyère, neufchâtel, parmesan,
roquefort), coffee (java), liquors (armagnac, bourbon [from Bourbon
County, Kentucky], tequila, etc.),4) and wines, e.g. (wine of) Bordeaux,

4) Gin [1714] is often cited in this context as deriving from Geneva. But gin is
produced by distilling rye or other grains with juniper berries, and is clipped
154 D. Gary Miller

Chablis, Champagne, etc. Sherry [1608] is backformed from sherris

[1597] ‘(wine) of Xeres’, a city in Spain (Mod. Jerez), where it was
produced. (Xeres in turn goes back to L Caesaris [urbs] ‘Caesar’s
[city]’.) A good example of the interchange between proper and common
nouns is madeira (Nyrop 1913 §489). L māteria (for māteriēs) ‘building
material; timber’ became madeira in Portuguese, which continued to
mean ‘material (timber) for construction’. The word was applied to a
small island whose forests furnished building material, resulting in the
name Madeira. Madeira (wine) [1598] in turn gave rise to the common
noun madeira.
Non-edible include a variety of diverse products. China(-ware)
remains transparent. From India come cashmere [1822] (Kashmir), calico
[a1505] (Calicut), madras (Madras), dungaree [1673], etc. Cologne was
first made in Cologne, on the Rhine. Jerseys were originally woolen
sweaters peculiar to fishermen on the Channel Island of Jersey. Suede is
extracted from suède gloves, a partial translation of F gants de suède
‘gloves of Sweden’ (Suède ‘Sweden’). Denim is clipped from sergé de
Nîmes ‘twilled cloth of worsted (for suits) from Nîmes, France. Worsted ’
firm-textured woolen yarn or fabric made from it’ is itself a metonymy.
It was by origin a product of Worthstede (now Worstead), a village in
Norfolk, England.

Ⅹ.. Pure Metonymy

So-called “pure” metonymy involves the transfer from one property to

another by one or more associated social contexts. It is perhaps better
referred to as semantic reanalysis (so Eckardt, Forthcoming, who does
not include it as a form of metonymy).

from geneva [1706] < Du. jenever < MDu geniver < OF gene(i)vre < L
jūniperus JUNIPER (HFW 178).
 Semantic Change 155

Social change can entail semantic reanalysis. Purchase (< OF

po(u)rchacier ‘ to pursue’ < VL prō + *captiāre, reformed from L
captāre ‘to strive to capture’) originally involved chasing and capturing.
When the same items came to be obtained in a more civilized manner,
purchase was reanalyzed as effected by money rather than capture (cf.
Copley 1981: 131).
L persōna ‘player’s mask; stage character’ was reanalyzed (already in
Latin) as the PERSON behind the character (Copley 1961: 120f.).
VL *dom(i)niārium ‘authority’ (cf. L dominium ‘sovereignty’) > OF
dangier > ME daunger ‘power; damage’ > DANGER, reanalyzed from
contexts in which people are endangered by those in their power (Stern
1931: 369).
Saboteur (F ‘one who shoes’) and sabotage, literally an act of
‘shoeing’, came to be associated with the throwing of a sabot (wooden
shoe) into a machine and breaking it. From there, the meaning was
generalized to describe any willfully destructive act.
Premises ‘real estate’ originally meant ‘the aforesaid, previously
mentioned’. In announcements and sales documents the details were
initially enumerated, viz. “house, with garden, land, tenements, etc.”
Later in the document, the premises was substituted for the description.
From this context, in which the referent was identical, the premises
was reanalyzed as the property itself (Stern 1931: 358ff.; Eckardt,
Forthcoming).
Arrant was originally a variant of errant (L errāns/errant-)
‘wandering, drifting; vagrant’. In contexts such as arrant thief ‘roving
robber’, arrant was reanalyzed as an intensive ‘notorious/manifest/
unmitigated’ and extended to traitor, rebel, coward, etc. After 1575 it
was “widely used as an opprobrious intensive, with fool, dunce, ass,
idiot ...” (Stern 1931: 394f.).
Rubric goes back to L rubrīca ‘red ochre (esp. as a pigment)’ (cf. L
ruber ‘red’), which was applied in the post-Augustan period to the
156 D. Gary Miller

chapter headings in red letters in books of law. The heading function

came to predominate, “and the word could then be extended to denote
other headings, not printed in red” (Stern 1931: 405).

ⅩⅠ. Part-whole Transfer (Synecdoche)

A part is substituted for the whole, i.e. the meaning of the whole is
transferred to a word that originally designated only a part of the
whole.5) At issue here is perception in terms of Gestalts: one perceives
the whole as an inseparable extension of the part. These substitutions
can be characterized as “context-independent conceptual associations”
(Eckardt, Forthcoming).
Examples include blade ‘sword’, head (count) [1535] ‘(number of)
people’, hands [1655] ‘laborers, workers’, brain [1914] ‘intelligent person’,
threads [1926] ‘clothes’, prick [1929] ‘despicable person’, wheels [1959]
‘car’.
There are several subclasses of part-whole substitutions. One
involves characterization by a garment or equipment, e.g. ensign,
redcoat, Green Beret, skirt [1914] ‘woman’, jock [1963] ‘versatile young
athlete’. More generally, this subclass has been termed container for
contained, e.g. the city for its inhabitants, the purse for money, the
bottle for alcoholic beverage, dish, paperback, etc. (cf. Bain 1893: 188f.).
The endless taxonomies (cf. Stern 1931: 319-82; Kövecses 2002:
152-6) can be avoided by mentioning several frequent types of
substitution: material for object (box for an item made of boxwood;
horn, iron, glass, rubber, etc.), object for person (hunk, sponge,
mouth-piece), and producer for product: a [painting by] Rembrandt,
Picasso, etc.; a [work by] Molière, Shakespeare (cf. Nyrop 1913 §297;

5) The inverse (whole for a part) belongs with transferred epithet, e.g. wearing a
mink [coat].
 Semantic Change 157

Stern 1931: 373; Kövecses 2002: 143ff.). These might also be considered
elliptical and/or more like transferred epithet (Ullmann 1957: 243f.),
illustrating the problem of subclassifying metonymy.6) The alternation of
action, agent, instrument, and location words has been mentioned (§2).
Historically, OE tūn ‘hedge; fence; enclosure’ (= Germ. Zaun ‘fence’)
came to designate the TOWN circumscribed by the enclosing item.
Similar examples have yielded many toponyms (G tópos ‘place’), e.g.
Hedges, Townsend, Newcastle, as well as common words. Traditionally,
the labyrinth is derived from G lábrus ‘double-bladed axe’ + -inthos
‘place (of)’, but more likely contained Anatolian Labarna (a Hattic king,
hence a regal tile) and meant ‘house of the king’ or the like (Furnée
1972: 397f.).
The part-whole relationship extends to entire situations, in which the
effect can be substituted for the cause, as in a reference to a program
as a yawn [1974], or to a stingy person as tight(-fisted).
So-called PROPRIETARY TERMS evolve by another type of
synecdoche, involving genericization, or promotion of one member of a
class to designate the class as a whole. Typical examples involve
kleenex (at the expense of tissue and the other companies that make
tissues), coke for cola (the leading representative of the class becomes
the name for the class), xerox for photocopy. Other proprietary terms
are aspirin (originally a brand name for a synthetic form of
acetylsalicylic acid), bandaid (Band-Aid), cellophane, jello (Jell-O),
thermos, etc.
This process is not unidirectional; there is also degenericization, as in
the case of E man, the original generic, which supplanted OE wer

6) With reference to the frequently discussed example, “The ham sandwich wants
a side dish of salad,” Kövecses (2002: 156) makes the following comment:
“The conceptual relationship might be specified as one of possession,
part-whole, or control, but none of them seems to fully capture the ‘essence’
of the kind of ‘contiguity’ that we feel holds between a customer and his or
her dish.”
158 D. Gary Miller

(werewolf, wergeld, etc.) as the male of the species. In this case the
male and the species itself were identified as coterminous.
The boundary between metaphor and metonymy is not always clear.
Emotion metaphors, for instance, generally “have a metonymic basis”
(Niemeier 2000: 198). Thus, heart is metonymic for the whole person, as
shown by expressions such as big-hearted, soft/tender-hearted,
hard-hearted, heavy-hearted, heartless, etc. (ibid., pp. 199ff); cf. gutless,
gutsy. In expressions like heart of stone, steel, etc., the materials are
“metaphorically mapped onto the domain of the heart” (ibid., p. 201).
These metaphors are in turn used metonymically: a personal trait
“stands for the whole moral outfit of that person” (Niemeier, p. 202).
Such interaction between metaphor and metonymy has been referred
to as METAPHTONYMY (Goossens 1990).
Examples like egghead involve complex interactions. First, there is
the metaphoric comparison of the head to an egg. Then, transferred
epithet imparts the meaning of ‘individual/person’ to the construct. The
further metonymic association of baldness (more directly similar to an
egg) with intellect (Tournier 1985: 294) is no longer present in the
meaning of egghead, which can apply to someone with very thick hair.
Similarly, redneck can apply to anyone with a comparable belief system,
bigwigs no longer wear wigs at all, but skinhead is still frequently
restricted to neo-Nazi types with shaved heads.

ⅩⅡ. Transfer of Characteristic (Antonomasia, Eponymy)

The transfer is from a proper noun to a common object or concept.7)

7) Eponyms from personal names can be separated from those based on

geographical, literary, mythic, and commercial names (Stockwell and Minkova
2001: 15-8). Our heuristic separates English from non-English eponyms. Most
of the examples cited here (plus others) are found in Boycott (1982) and
 Semantic Change 159

Procrustes, in Greek legend, forced overnight travelers to fit his iron

bed by stretching them or cutting them down to size, whence
procrustean, of the use of violence to force uniformity or conformity.
Tantalus, mythical king of Phrygia, abused the privilege of dining
with the gods either by sharing ambrosia with mortals or by serving
his son Pelops to the gods for dinner. As a punishment, he was doomed
to be forever in sight of fruit and water that receded whenever he
reached for it, hence tantalize.
Pandarus, an archer who wounded Menelaus in Homer’s Iliad, served
in Chaucer’s Troilus & Criseyde as the intermediary between the two
lovers, Priam’s son Troilus, and Cressida, daughter of a priest who fled
Troy leaving her there, whence pander ‘liaison in sexual intrigues; pimp;
procurer’.
Stentor, a Greek herald in the Trojan War known for his voice of
fifty men (Iliad 5.785), is the source of a stentorian (very loud) voice.
Epicurus [c.341-c.270] started a school (way of life) in Athens [-306],
advocating atheism, high living, and sensual gratification, and forbidding
marriage, children, and participation in public life. From the associations
with his name is derived E epicure(an), one who cultivates refined
tastes, especially in food and wine.
Pyrrhus [319/8-272] lost so many men in his defeat of the Romans at
Asculum [-279] that his name gave rise to the pyrrhic victory.
Philippic denotes a bitter invective, as used by Cicero of his
harangues against Marc Antony, modeled after Domosthenes’
denouncement of Philip Ⅱ of Macedon, the father of Alexander the
Great.
Quintus Roscius [c-1], a friend of Cicero’s, was such a brilliant comic
actor that roscian denotes eminence in the performing arts.
Caesar, a name acquired by Julius’ grandnephew Octavian (known
today by his honorary title Augustus conferred in -27) after his

Freeman (1977).
160 D. Gary Miller

adoption by Julius Caesar. Since Octavian/Augustus was the first

Roman emperor, the title Caesar was adopted by subsequent emperors,
and came by association to be synonymous with imperātor ‘emperor’,
giving rise to Germ. Kaiser, Slavic czar, tsar, etc.
L Monēta ‘she who advises’ (?) (from Etruscan?), a cult title of Juno,
was used of her temple at Rome where money was coined, then of any
mint, then ‘coinage, money’ [first in Ovid], yielding F monnaie, E money.
John Duns Scotus [?1265-1308], born in Duns, Scotland, was a
scholastic philosopher and theological archconservative. His followers,
Duns men or Dunses [a1530], because of their hair-splitting, were
viewed contemptuously by opponents and later philosophers, whence E
dunce (men) [a1590].
German Taler ‘valleyer’, clipped from Joachimstaler ‘one from
Joachimsthal’ (Joachim’s Dale), a place in Bohemia where silver was
minted in c16, gives E dollar.
Béchamel sauce is named after its inventor, Louis de Béchamel,
steward of Louis ⅪV [1643-1715].
Chauvinism is eponymous for the overzealous French patriot Nicolas
Chauvin [fl. 1815] who continually sang Napoleon’s praises.
Ethnic names tend to evolve into pejoratives, perhaps the best known
of which are Viking/viking and Vandal/vandal (cf. Nyrop 1913: 386f.).
The sciences have many well-known eponyms, e.g. ampere, (deci)bel
(< Bell), hertz, ohm, volt, watt.

ⅩⅢ. English-Specific Eponymy

The English language is fond of remembering people eponymically.

Many common objects bear noble names. The cardigan sweater is
named after James Thomas Brudenell [1797-1868] who became the
Seventh Earl of Cardigan. He so frequently wore the colorless sweater
 Semantic Change 161

with buttons down the front that it came to be associated with him.
Dr. Thomans Bowdler [1757-1825] published a ten-volume edition of
The family Shakespeare (Bath, 1818) with all “indelicate” passages
excised. The word bowdlerize was coined in 1836 for this type of moral
censorship.
The Fourth Earl of Sandwich (John Montagu) [1718-92] was a
passionate gambler. To avoid leaving the gambling table, he ordered that
slices of bread with roast beef between them be brought to him,
inventing the sandwich [1762]. His ineptitude as First Lord of the
Admiralty contributed to British defeat in the American Revolution and
possibly also to the American fondness for sandwiches.
Commoners so immortalized include John L. McAdam [1756-1836], a
Scottish engineer who developed macadam [1824], and Charles
Macintosh [1766-1843], a Scottish chemist who invented the
mac(k)intosh [1835] rubberized cloth raincoat.
Sir Thomas Crapper [1837-1910], Chelsea Sanitary Engineer, in 1884
invented a modern toilet, Crapper’s Valveless Water-Waste preventer,
and Crapper’s Seat Action Automatic Flush. He has been awarded the
dubious accolade of several slang clippings.8)
Shyster may be from Scheuster, an unscrupulous New York attorney
of the 1840s.
Blurb [1907] is named for a fictional character, Belinda Blurb, used as
an advertising gimmick on the dust jacket of 500 copies of Gelett
Burgess’ Are You a Bromide? [1906].
Bikini [1946] received its name from the type of swimsuit worn by
women on the Bikini atoll of the Marshall Islands.
Clipping frequently obscures metonyms from proper names, e.g. Saint
Audrey (pronounced /sen.tɔ.drɩ/ in British) > tawdry ‘cheap and flashy;

8) Hence the slang uses ‘residue; dregs; scrap’ of E crap [LME] < AF crappe
‘chaff; husk’ (Rothwell 1996: 47f.); cf. ON krap ‘slush’; Ice, Norw. krap ‘small
clumps of ice or snow’ (de Vries 1977: 328f.).
162 D. Gary Miller

vulgar’, from the cheap, gaudy lace neckties sold at fairs in honor of St.
Audrey [†679].
St. Mary of Bethlehem [bεþlǝm], a London insane asylum > [bεðlǝm]
> bedlam ‘uproar, total confusion’.

ⅩⅣ. Folk Etymology

The nature of folk etymology is that some X is reanalyzed as the

nearest Y it “sounds like” and that also in some manner accounts for
the meaning. More formally, folk etymology involves hyponymy:9) “the
affecting form designates a superordinate term, and the affected one
comes to resemble the superodinate term formally and denote a subtype
semantically” (McMahon 1994: 184). This, of course, presupposes that
words are stored and processed according to sound and meaning, and
that the relationship is one of analogy (Coates 1987). In the words of
Meier (1989: 71), “‘wrong’ etymologies must be considered as a
systematic expression of linguistic psychology, because the kind of
equation involved is a creative activity endeavouring to make a word
significant again ― even where the new significance is only apparent”
(cf. Rundblad and Kronenfeld 2000).
Folk etymology is frequent in the adaptation of foreign expressions
(Ullmann 1957: 91), e.g. mayday for F m’aider ‘help me’, but also applies
to well-known words, sometimes in surprising ways: “A folk etymology
like sparrow-grass for asparagus ... borders on the lunatic” (Sihler 2000:
87).
Popular attempts at etymologizing frequently leave a portion of the
word unanalyzed. For instance, the cray part of crayfish (F écrévisse)
remains unmotivated. The twisting of F femelle ‘little woman’ into E

9) In hyponymy, the meaning of one word is included in another, e.g. affirm,

deny, and narrate include the meaning of say.
 Semantic Change 163

female fit the cultural notion of the extraction of the female from the
male, and was paralleled in woman, even though both formations leave
an unexplained residue (fe-, wo-), which have prompted recent attempts
to resolve (see Rundblad and Kronenfeld, pp. 26f.).
Examples include:10)
a) hurricane ‘tropical cyclone’ < Sp. huracán (cf. Port. furacão) <
Carib huracán, furacan, West Indian (Taino) hurakan ‘an evil spirit of
the sea; hurricane’. The borrowed form was altered to match hurry (cf.
Shipley 1984): 185), an existing word with a meaning consistent with
the storm’s velocity.
b) miniature [1586] was borrowed from French, in turn from Ital.
miniatura ‘painting’ (esp. the illuminations in Medieval manuscripts), a
derivative of miniare ‘to illuminate’ (L miniāre ‘to color red with
cinnabar [minium MINIUM])’. Under the influence of minus, minuscule,
etc., and since the illuminations were generally small, miniature was
reanalyzed as ‘small picture or decorative letter’, then ‘on a small or
greatly reduced scale’ (cf. Nyrop 1913: 327f.).
c) sacri-leg-ious / sacrilege ‘desecration’, from L sacri-leg-us ‘one
who steals sacred things’ (leg- ‘gather; pluck; steal’), tends to be
reanalyzed as if sac-relig-ious by analogy with religion, religious, in
the same semantic field. Religious itself in the Christian sense may
result from a folk etymology. Cicero (On the Nature of the Gods
2.28.72) derived L religiō from re-legere ‘re-collect’ and understood it
as a scruple that hinders (a subjective restraint). The Christian writers

10) The alleged example dormouse (pl. -mice), a squirrellike Old World rodent of
the family Gliridae, is problematic. Because it is nocturnal and a hibernator,
often found sleeping, the animal’s name was supposedly derived from L
*dorm-ōsus ‘full of sleep’, which yielded ME dormowse. From the similarity to
mouse, another rodent, the subsequent history was influenced by forms of
mouse, including its irregular plural mice. Another suggested etymology is F
dormeuse (fem.) ‘sleeper’; or English dialect dorm ‘sleep’ (< F dormir ‘to
sleep’) + mowse ‘mouse’ (cf. Partridge 1983: 164).
164 D. Gary Miller

Lactantius and Tertullian derived it from religāre ‘tie back’ and

understood it as a binding obligation, an objective tie that ‘binds’ us to
the deity. Benveniste (1969: ii. 267-73) argues for the former etymology;
Watkins (2000: 47) accepts the latter (*leig-1 ‘bind’); Hoad (CDEE 397)
leaves it open.
d) hero (sandwich), another term for a sub(marine) sandwich or
hoagie, derives ultimately from G gūros ‘circle’, gūrós ‘round’ (cf. gyro
[a1975]). In Modern Greek, it is used of a sandwich made from meat
grilled on a turning spit and pronounced /ƛiro/ which, of course sounds
like Eng. hero. The meaning is somewhat remote but possibly
rationalized as a reference to the size/stalwortness of both the sandwich
and its consumer.
e) cutlass is etymologically from OF coutelas, a derivative of coutel
‘knife’ (< L cultellus, diminutive of culter ‘knife’). The word thus had
nothing to do with E cut etymologically but, functionally, cutting is
what it does, so modification of the borrowing to conform to the
spelling of cut is hardly unexpected.
f) Aphrodite (G Aphrodítē) is of unknown origin (DELG 148). What
is generally admitted is that it must be from a pre-Greek source and
that the Greek connection with aphrós ‘foam’, whence the myth that she
was born from the foam of the sea, is a folk etymology (cf. Nyrop
1913: 331).
g) andiron was borrowed from OF andier (modern landier) ‘firedog’.
The ending was identified with ME yre, yren ‘iron’, whence the modern
spelling andiron (Stern 1931: 234).
A word of caution. One must be careful not to confuse real
etymologies (X derives from Y) with folk etymologies, according to
which X is transformed into or reshaped after Y, where at least part of
Y is a previously existing word that somehow resembles X in form
and/or meaning.
 Semantic Change 165

ⅩⅤ. Tabu Substitution and Deformation

1. Religious Tabu

A Polynesian (Tongan) word brought back by Capt. Cook in 1777,

taboo (in technical writings generally spelled tabu) originally pertained to
the sacred and prohibited, the profane [1483], in the etymological sense
of L profānus (prō ‘before’ ＋ fānum ‘temple’) ‘contemptuous of sacred
things; sacrilegious; secular’. Any word formation or semantic process
should in principle be able to be used to construct euphemisms (cf.
Warren 1992; Ham 2001).
In most societies it is inappropriate to utter words of religious import
with contempt. In the language of swearing, it is therefore not
surprising to find many phonemic modifications (Warren 1992) of E God,
e.g. gog [1350s], cod [1569], gad [1611], gadzooks (= God’s hooks)
[1650s], egad [1673], gosh [1743], golly [1743], drat (God rot) [1844],
doggone (God-damn) [1851], Godfrey [1853], golda/urn(ed) [1856], good
grief [1900, etc. (cf. Hughes 2000: 47).11) Similarly, for Christ, one finds
such deformations as cripes [1910] (< by Christ’s stripes), and for
Jesus: gee [1895], jeepers [1929], jeez(e) [1970]. The devils gets at
least twenty-five euphemisms. For hell, there are such circumlocutions
as Sam Hill [a1840], but the main tabu form is heck [a1855]. Damn(ed)
gets deformed into deuced [a1785], tarnation [a1785], darn [1789], dashed
[1881], dang(ed) [1886], etc.
One of the semantic devices of tabu replacement is ANTIPHRASIS
(G antíphrasis ‘opposite phrasing’), the use of words of positive sense in
place of those with negative connotations. The Greeks replaced Erīnúes

11) Some French tabu forms for Dieu ‘God’ are known in English: Ventre Dieu
‘God’s belly’ → ventrebleu; Chair Dieu ‘God’s flesh’ → Charbieu (Villon),
Corps Dieu ‘God’s body’ → Corbieu (Rabelais), etc. (Nyrop 1913: 271f.).
166 D. Gary Miller

[Myc.＋] ‘Erinyes; the Furies’, the avenging deities (goddesses who fulfil
curses and avenge crimes), with Eumenídes [c-6/5] [‘well-minded
(ones)’] ‘Eumenides (the gracious goddesses)’. Another group of
goddesses, the Moĩrai (pl. of G moĩra ‘portion, share; lot; fate; destiny’;
i.e. ‘Destiny’ personified) = L Fāta ‘Fates’, were identified in Latin with
the Parcae, birth spirits, goddesses of birth (*perǝ-1 ‘to produce’). The
identification was enhanced by the Roman folk etymology which
connected Parcae with the verb parcere ‘to spare’, but, beginning with
Horace, the Roman tradition loantranslated the Greek antiphrasis
Khárites as Grātiae ‘Graces’, the goddesses who confer all grace,
including the favor of victory in the games.
Whether or not the Samniam town Mal(e)ventum was derived from
G Malóenta (ACC) ‘apple orchard’, the Romans interpreted it as ‘illcome’
and changed the name to Beneventum ‘wellcome’ (cf. Löfstedt 1959:
185.).
When there is a general prohibition against even pronouncing the
word, its true form may cease to be known, as happened with the
Hebrew word for ‘god’, YHWH, for which various guesses have been
ventured as to the vowels, e.g. Jehovah, Yahweh. Similarly, the name of
the original head of the Indo-European pantheon cannot be
reconstructed. Some of the variants are ―

Greek (terpi-)kéraunos ‘(enjoying the) thunder-bolt’ (epithet of Zeus)

Lithuanian perkūnas ‘thunder (god); lightning bolt’ (*perkw-ūno-)
Slavic perun- ‘thunder god’ (Perunov dub [1302] ‘Perun’s oak’)
Sanskrit parjánya- ‘rain god; thunder god’ (*pergw-én-yo-)
Gothic fairguni ‘mountain’
Old Norse Fjǫrgyn ‘Earth’ (mother of Þórr, god of thunder)
(w)
Celtic-Latin Hercynia, a mountain-forest region (*perk -un-yā)
Latin quercus ‘oak tree’ (*perkw-u-)
 Semantic Change 167

It defies common sense to imagine that these words are not related in
some manner, and yet it is impossible to reconstruct a common form
from which they can all be derived (Meillet 1965: 333; cf. de Vries 1977:
126; Delamarre 2001: 139).
Certain animals are subject to religious tabu (Emeneau 1948). The
Indo-European word for ‘bear’ *h2ŗtḱ-o- (G árktos, L ursus, Hittite
ḫartaggaš) is preserved only in areas where there were no bears
(Meillet 1906). Elsewhere, it was replaced, like E bear (= Lithuanian
béras ‘brown’), OIr. milchobur [‘honey-desiring’] (Watkins 1962: 114ff.;
Uhlich 2002: 418f.), Old Slavic medvĕdĭ (cf. Skt. madh(u)v-ád-
‘honey-eater’). In Celtic, the Indo-European word is preserved only in
names, e.g. Art, Hart (OIr. ó h-art ‘son of bear’), Welsh arth in Arthur
(Welsh arth-gwyr ‘bear-man’). The bear is an animal to dread, to hunt,
something mystical. It is dangerous to say its name unless it is
associated with a person, hence the preservation in names, where the
original meaning is eventually forgotten. Each of the northern languages,
where bears are prevalent, has replaced the original word with a
euphemism.

2. Secular Tabu

Not all tabu words are religious, of course, even though it is not
always easy to draw the line. What is religious in one culture may be
secular in another, both being a reflection of cultural ideology (cf. Ham
2001). Tabus surrounding natural life functions (birth, death, men-
struation, evacuation, sex(uality), pregnancy, procreation) have been
particularly rampant. Sickness is also prone to tabu substitution. For L
aeger ‘sick’, the Romans frequently substituted īnfirmus INFIRM, gravis
GRAVE, languidus LANGUID, and Vulgar Latin negated habitus ‘in
good shape or condition’as male habitus [c1: Sabinus], whence F malade
‘sick’; cf. MALADY (Nyrop 1913: 278f.). This use of LITOTES
(negation of the positive term as a euphemism for the negative) is very
168 D. Gary Miller

common. Nyrop (1913 §443) cites maladroit, dishonest, disagreable,

inactive, incapable, inclement, incompetent, incontinent, indecent,
indelicate, indiscreet, unjust, etc.
Drunkenness is subject to humorous euphemisms, e.g, tipsy, lit up.
Mental retardation has been the source of frequent mockery and
euphemism: imbecile (L imbēcillus ‘physically weak; feeble; fragile’),
idiot (F < L idiōta ‘layman; amateur’ < G idiṓtēs ‘private person;
common person; unskilled or ignorant person’ [Nyroop 1913 §§180, 184,
397]), cretin (< F crétin ‘idiot’ < Swiss F crestin ‘Christian’ (L
Christiānus), the idea being that such an individual is deformed but still
human [Nyrop 1913: 284f.]).
Certain body parts are subject to tabu; cf. butt(ocks), rear (end),
posterior, derrière, gluteus maximus. Among bodily functions, there are
such euphemistic deformations as shucks, shoot [1880] for shit, and even
dipstick [1963] for dipshit [1962].
For fucking [a1890], one finds frigging [c.1785], flipping [1911],
mucking [1929] (but muck around [1856], long before fuck around
[1922]), effing [c.1929], fricking [1936], and, two that were popular among
my friends in the 1950s, freakin’, frappin’, and, still more recently,
shtupping [a1970], bonking [a1986]. Finally, there is the f-word [1988]
(Ayto 1990: 147). In all, English has some 800 words for copulation,
1200 for vagina, 1000 for penis, and 2000 for whore (Allen and Burridge
1991: 96). Ham (2001) argues that sex has become a great secret,
“necessitating its constant discussion and the subsequent creation of
new euphemisms” (§2.1). Writers through history have been concerned
with politeness “either by adhering to the historical and cultural
expectations, or by deliberately shattering them” (Ham, §5.1).
Euphemistic adjectives suggest more colorful terms, e.g. blasted
[1843], blooming [1885], blankety-blank [1888], and more recently, ideo-
grams(!*#&*%!). Ham (2001, §5.2) also mentions graphological dele-
tion(****), phonemic deletion (Did you ―?” for “Did you have sexual
 Semantic Change 169

intercourse?”), and aural deletion (radio and TV “bleeps”).

Even the facilities associated with bodily functions are prone to
euphemistic replacement in so-called polite circles, e.g. the john [a1375],
closet of ease [1662], water closet [1775] (> W.C.), lavatory [1845]
(earlier as ‘bath’ [a1375] and ‘bathing room’ [a1670], potty [1850],
commode [1851], toilet [1885] ([a1670] as ‘toiletry’), bathroom, etc. (cf.
Williams 1986: 186f.). Many of these are borrowed, as is frequent with
euphemisms (Warren 1992); cf. Cicero’s many Greek euphemisms (Nyrop
1913: 265).
A word homophonous to a tabooed word can be replaced. When arse
came to rhyme with ass, ass was replaced by donkey. When cock
became a common word for ‘penis’, the cock (bird) was replaced by
rooster (Bloomfield 1933: 396; de la Cruz Cabanillas and Tejedor
Martinez 2002: 238ff.). Since gay has replaced homosexual, it is avoided
as a synonym for ‘happy’ (cf. Ham 2001, §2.0). Pussy is rarely heard in
the context of cats. Such displacements are not automatic, of course.
Lass (1997) cites examples of enduring homophony, such as dike, come,
boob, crap, clap, cream, shaft, screw, suck. On balance, not all of these
are used in the same contexts (on this criterion, see Ullmann 1957:
132ff.). Also, horny, cited by Lass, is becoming rare, at least in
American English, in the sense of ‘having horns’ since the sexual sense
has come to predominate.

XⅥ. Loan Translations and Other Contact Phenomena

1. Calques

In calquing, a native equivalent is substituted for each constituent of

a semantically transparent foreign word. Superman is a calque on Germ.
Übermensch (über ‘over’ plus Mensch ‘human being’). Other German
calques: world-famous (weltberühmt), antibody (Antikörper), standpoint
170 D. Gary Miller

(Standpunkt), daydream (Tagtraum), etc. (Tournier (1985: 317)).

Rather than borrow L ēvangelium (G eu- ‘good’ ＋ angel- ‘message’),
Old English loantranslated it as gōdspell [good-message] (> GOSPEL).
Likewise, L omnipotēns OMNIPOTENT was loantranslated into Old
English as ealmihtig (> ALMIGHTY), and OE hēþendōm (>
HEATHENDOM) is a calque on LL pāgānismus (PAGANISM).
The Homeric nymph Calypso (G Kalupsṓ) is derived from G
kalúpt-ein ‘to hide’ [*kel-2] (see DELG 488; Sihler 1995 §322), meaning
something like ‘the Concealer; she who conceals’. Since antiquity, she
was associated with the island of Malta (Phoeniciam Malet(h) ‘shelter,
haven; hiding place’), implying a connection via loan translation.

2. Semantic Transfer

Exposition is used in the sense of ‘exhibition’ under French influence

(Stern 1931: 222). Novel was borrowed from French as ‘something new;
novelty’ [1460] and ‘news, tidings’ [1475]. In 1566, it acquired the sense
of ‘short story’ under the influence of Ital. novella, used of the stories in
Boccaccio’s Decamerone (Stern 1931: 223).
L conclūsiō ‘the enclosing (of an area)’ was extended to
CONCLUSION under the influence of G sumpérasma ‘conclusion of a
syllogism’ (Coleman 1989: 83).
Hebrew ml’k ‘messenger’ also meant ‘angel’. Since Greek had no
equivalent, the Bible redactors loantranslated ml’k as ángelos ‘messenger’
with the extended sense of ‘angel’ (Meillet 1951 [1925]: 40f.; cf.
Geeraerts 1997: 89, 101).
G ptõsis ‘a fall(ing)’ was also used for ‘(grammatical) case’. Varro
loantranslated the term as L cāsus ‘a fall(ing)’, thereby extending the
meaning to CASE (cf. Ullmann 1962: 167). The names of the cases
were also modeled on the Greek (Wackernagel 1926: 13-20; Coleman
1989: 83f.).
 Semantic Change 171

OE eorl ‘free man; noble’, during the reign of Cnut [?994-1035],

acquired the sense of the Nordic cognate jarl ‘viceroy; provincial
governor’ (see Pons Sanz 2003). In the Anglo-French period, it was
used for F comte COUNT, continued in MnE earl (Meillet 1965
[1905-6]: 249).

3. Calque Coinage

A word may be created from native constituents as a calque on a

foreign word. Cicero (Academica 1.25) coined L quālitās ‘characteristic;
QUALITY’ from L quālis/quāli- ‘of what sort’ ＋ －(i)tās ‘－ness’ on
the model of G poiótēs ‘quality’ from poĩos/poio－ ‘of what sort’ ＋ -tēs
‘ness’ (cf. Coleman 1989: 80).
From the possible but non-existing participle stem essent- ‘being’
plus -ia, L essentia ESSENCE was coined after G ousíā ‘essence’ (built
on the participle stem ont- ‘being’ + abstract -íā). Seneca attributes
essentia to Cicero but Quintilian attributes it to the rhetorician Verginius
Flavus or the philosopher Sergius Plautus (cf. Coleman 1989: 80f.).

4. Calque Remodeling

Under the influence of OF bien venu (lit. ‘well come’), OE wilcuma

(lit. ‘pleasing-comer’ with wil(l) ‘pleasure’) ‘guest’ was reshaped to
welcome with well plus the past participle of come (cf. Stern 1931: 234).
Indonesian asbut ‘smog’ is a blend calqued on the constituents of E
smog: asap ‘smoke’ x kabut ‘fog’ (Sihler 2000: 130).

ⅩⅦ. Conclusion

Perhaps nowhere in historical linguistics is the essentially human

character of language as evident as in semantic change. The main
172 D. Gary Miller

factors involved in semantic change are expressivity and cognition. A

creative metaphor simultaneously reflects our mental interconnections
and expands our schema for a given concept beyond the prototypical.
The role of cognition is seen repeatedly in changes from the objective
to the subjective and/or evaluative, the tendency to personalize referents
to real world items. And extreme form of personalization is encountered
in tabu substitutions and deformations. At a more basic level, however,
is the implicit belief (or hope) that if we can get a GRIP on the
universe, and potentially manipulate our environment, we have the
capacity (albeit illusory) of understanding it better. At the same time,
things we can manipulate can evolve into less manipulable abstractions,
illustrating the human countercapacity to distance itself from the
environment.
Our desire to analyze things and events in terms that make sense to
us is evident in folk etymologies as well as in various forms of
metonymy and semantic reanalysis. Substitutions of material for object,
producer for product, etc., reflect and implicit desire to reduce the
vastness of the universe by conflating wherever possible. They are also
economy moves to abbreviate language for production efficiency. At the
same time, like all semantic changes, they provide valuable information
on the natural associations in our mind/brain.
 Semantic Change 173

Dating Conventions
[c1] first century AD/CE [319-270] 319 BC(E) to 270 BC(E)
b
[c1 ] beginning of c1 [a1845] before 1845
e
[c1 ] end of c1 [c.1845] about 1845
m
[c-1] first century BC/BCE [fl.c1 ] flourished middle of c1
[c-6/5] sixth or fifth century BC(E) [†679] died in 679

Abbreviations
ACC accusative MDu Middle Dutch
AF Anglo-French ME Middle English
Arab. Arabic ML Medieval Latin
Ch. Chaucer MnE Modern English
Du. Dutch Myc. Mycenaean (Greek)
E English Norw. Norwegian
F French OE Old English
f. (one page) following OF Old French
ff. (two pages) following OIr. Old Irish
G (Ancient) Greek ON Old Norse
Germ. German Pers. Persian
Gmc. Germanic pl. plural
ibid. in the same reference Port. Portuguese
Ice Icelandic PPP past passive participle
IE Indo-European Skt. Sanskrit
Ital. Italian Sp. Spanish
L Latin VL Vulgar Latin
LL Late Latin w.lit with literature (reference)
LME Late Middle English

Bibliographical Abbreviations
CDEE = The Concise Oxford Dictionary of English Etymology, ed.
Terry F. Hoad. Oxford: Oxford University Press (1996 [1986]).
174 D. Gary Miller

DELG = Dictionnaire étymologique de la langue grecque. By Pierre

Chantraine. Paris: Klincksieck (1968-80)
DELL = Dictionnaire étymologique de la langue latine. By Alfred
Ernout and Antoine Meillet. 3rd edn. 2 vols. Paris: Klincksieck
(1951).
HFW = A History of Foreign Words in English. By Mary Sidney
Serjeantson. New York: Barnes & Noble (1961 [1935]).
ODS = The Oxford Dictionary of Slang, ed. John Ayto. Oxford: Oxford
University Press (1998).
 Semantic Change 175

Works Cited

Allen, K., and K. Burridge. 1991. Euphemism and Dysphemism:

Language Used as Shield and Weapon. Oxford: Oxford University
Press.
Ayto, John. 1989-90. The Longman Register of New Words. 2 vols.
Essex: Longman.
Bain, Alexander. 1893. English Composition and Rhetoric. Enlarged
Edition. Part First: Intellectual Elements of Style. London:
Longmans, Green & Co.
Barcelona, Antonio (ed.). 2000a. Metaphor and Metonymy at the
Crossroads: A Cognitive Perspective. Berlin: Mouton de Gruyter.
__________ 2000b. “Introduction: The Cognitive Theory of Metaphor and
Metonymy.” In Barcelona (2000a: 1-28).
Bauer, Laurie. 2002. “What You Can Do With Derivational Morphology.”
In Bendjaballah et al. (2002: 37-48).
Bendjaballar, Sabrina, Wolfgang U. Dressler, Oskar E. Pfeiffer, and
Maria D. Voeikova (eds.). 2002. Morphology 2000: Selected Papers
from the 9th Morphology Meeting, Vienna, 24-28 February 2000.
Amsterdam: John Benjamins.
Benveniste, Émile. 1949. “Euphémismes anciens et modernes.” Die
Sprache 1: 116-22. (Repr. in Benveniste 1966: 308-14.)
___________ 1966. Problèmes de linguistique générale. Paris: Gallimard.
___________ 1969. Le vocabulaire des institutions indo-européennes. 2
vols. Paris: Minuit.
Besch, Werner, Anne Better, Oskar Reichmann, and Stefan Sonderegger
(eds.). 1998. Sprachgeschichte: Ein Handbuch zur Geschichte der
deutschen Sprache und ihrer Erforschung. 2nd edn. Berlin: Walter
de Gruyter.
Bloomfield, Leonard. 1933. Language. New York: Holt, Rinehart &
Winston.
Boycott, Rosie. 1982. Batty, Bloomers and Boycott: A Little Etymology
of Eponymous Words. New York: Peter Bedrich (1983).
Coates, Richard. 1987. “Pragmatic Sources of Analogicl Reformation.”
Journal of Linguistics 23: 319-40.
176 D. Gary Miller

Coleman, Julie and Christian J. Kay (eds.). 2000. Lexicology, Semantics

and Lexicography: Selected Papers from the Fourth G. L. Brook
Symposium, Manchester, 1998. Amsterdam: John Benjamins.
Coleman, Robert. 1989. “The Formation of Specialized Vocabularies in
Philosophy, Grammar and Rhetoric: Winners and Losers.” In
Lavency and Longrée (1989: 77-89).
Copley, J. 1961. Shift of Meaning. London: Oxford University Press.
Dalton-Puffer, Christiane, and Nikolaus Ritt (eds.). 2000. Words:
Structure, Meaning, Function: A Festschrift for Dieter Kastovsky.
Berlin: Mouton de Gruyter.
De la Cruz Cabanillas, Isabel, and Cristina Tejedor Martinez. 2002. “The
Horse Family: On the Evolution of the Field and its Metaphorization
Process.” In Díaz Vera (2002: 229-54).
Delamarre, Xavier. 2001. Dictinnaire de la langue gauloise: Une
approche linguistique du vieux-celtique continental. Paris: Éditions
Errance.
Díaz Vera, Javier E. (ed.). 2002. A Changing World of Words: Studies
in English Historical Lexicography, Lexicology and Semantics.
Amsterdam: Rodopi.
Eckardt, Regine (Forthcoming). Meaning Changes by Semantic
Reanalysis: Implications for the Reconstruction of Language. Oxford:
Oxford University Press.
Emeneau, Murray B. 1948. “Taboos on Animal Names.” Language 24:
56-63.
Feyaerts, Kurt. 2000. “Refining the Inheritance Hypothesis: Interaction
Between Metaphoric and Metonymic Hierarchies.” In Barcelona
(2000a: 59-78).
Fischer, Andreas (ed.). 1989. The History and the Dialects of English:
Festschrift for Eduard Kolb. Heidelberg: Carl Winter.
Freeman, Morton S. 1997. A New Dictionary of Eponyms. Oxford:
Oxford University Press.
Fritz, Gerd. 1998a. Historische Semantik. Stuttgart: J. B. Metzler.
_______ 1998b. “Ansätze zu einer Theorie des Sprachwandels auf
lexikalischer Ebene.” In Besch et al. (1998: 860-74).
Fruyt, Michéle. 1989. “Étude sémantique des “diminutifs” latin: les
suffixes -ulus, -culus, ellus, -illus ... de-substantivaux et de-
 Semantic Change 177

adjectivaus.” In Lavency and Longrée (1989: 127-38).

Furnée, Edzard J. 1972. Die wichtigsten konsonantischen Erscheinungen
des Vorgriechischen: Mit einem Appendix über den Vokalismus.
The Hague: Mouton.
Geeraerts, Dirk. 1997. Diachronic Prototype Semantics: A Contribution
to Historical Lexicology. Oxford: Clarendon Press.
Goddard, Cliff. 1998. Semantic Analysis. A Practical Introduction.
Oxford: Oxford University Press.
Goossens, Louis. 1990. “Metaphotonymy: The Interaction of Metaphor
and Metonymy in Expressions for Linguistic Action.” Cognitive
Linguistics 1: 323-40.
Gordon, Arthur E. 1983. Illustrated Introduction to Latin Epigraphy.
Berkeley & Los Angeles: University of California Press.
Görlach, Manfred. 2000. “Conceptual and Semantic Change in the History
of English.” In Dalton-Puffer and Ritt (2000: 95-109).
Guiraud, Pierre. 1980. Le langage du corps. Paris: Presses Universitaires
de France.
Ham, Kerry. 2001. “The Linguistics of Euphemism: A Diachronic Study
of Euphemism Use and Formation.” MA thesis, University of
Surrey, UK.
Hecht, Max. 1888. Die griechische Bedeutungslehre: Eine Aufgabe der
klassischen Philologie. Leipzig: Teubner.
Hoad, Terry F. (ed.). 1996 [1986]. The Concise Oxford Dictionary of
English Etymology. Oxford: Oxford University Press.
Hogg, Richard M. and Linda van Bergen (eds.). 1998. Historical
Linguistics 1995: Selected Papers from the 12th International
Conference on Historical Linguistics, Manchester, August 1995. Vol.
ii: Germanic Linguistics. Amsterdam: John Benjamins.
Hooker, James T. 1980. Linear B: An introduction. Bristol: Bristol
Classical Press.
Hughes, Geoffrey. 2000. A History of English Words. Oxford: Blackwell.
Kastovsky, Dieter and Arthur Mettinger (eds.). 2001. Language Contact
in the History of English. Frankfurt am Main: Peter Lang.
Koivisto-Alanko, Päivi. 2000a. “Mechanisms of Semantic Change in
Nouns of Cognition: A General Model?” In Coleman and Kay (2000:
35-52).
178 D. Gary Miller

_________ 2000b. Abstract Words in Abstract Worlds: Directionality and

Prototypical Structure in the Semantic Change in English Nouns of
Cognition. Helsinki: Société Néophilologique.
Kövecses, Zoltán. 2002. Metaphor: A Practical Introduction. Oxford:
Oxford University Press.
Krahe, Hans, and Wolfgang Meid. 1967. Germanische Sprachwissen-
schaft. Vol. iii: Wortbildungslehre. (= Sammlung Göschen Band
1218b.) Berlin: Walter de Gruyter.
Lass, Roger. 1997. “Arse Longa, Vita Brevis: Last Words on “Harmful
Homophony”.” Studia Anglica Posnaniensia 32: 21-31.
Lavency, Marius, and Dominique Longrée (eds.). 1989. Proceedings of
the Vth Colloquium on Latin Linguistics (= CILL 15: 1-4). Louvain-
la-Neuve: Peeters.
Leisi, Ernst. 1974. Das heutige Englisch: Wesenzüge und Probleme.
Heidelberg: Carl Winter.
Liberman, Anatoly. 2000. “The Etymology of English boy, beacon, and
buoy.” American Journal of Germanic Linguistics & Literatures 12:
201-34.
Lipka, Leonhard. 2000. “Word-Formation and (Proper) Names: A
Neglected Field.” In Dalton-Puffer and Ritt (2000: 187-203).
Löfstedt, Einar. 1959. Late Latin. Cambridge, Mass.: Harvard University
Press.
Logopédies [no editors listed]. 1988. ΗΔΙΣΤΟΝ ΛΟΓΟΔΕΙΠΝΟΝ / Logopédies:
Mélanges de philogie et de linguistique grecques offerts à Jean
Taillardat. Paris: Peeters/Selaf.
McMahon, April M. S. 1994. Understanding Language Change.
Cambridge: Cambridge University Press.
Meier, Hans H. 1989. “Etymological Modes in Modern English.” In
Fischer (1989: 59-76).
Meillet, Antoine. 1905-6. “Comment les mots changent de sens.” In
Meillet (1965: 230-71).
_______ 1906. “Quelques hypothèses sur des interdictions de vocabulaire
dans les langues indo-européennes.” In Meillet (1965: 281-91).
_______ (n.d.). “La religion indo-européenne.” In Meillet (1965: 323-34).
1925. “Les interférences entre vocabulaires.” In Meillet (1951:
36-43).
 Semantic Change 179

1951-65. Linguistique historique et linguistique générale. Vol. ii

(1951). Vol. I (1965). Paris: Honoré Champion.
Miller, D. Gary. 1993. Complex Verb Formation. Amsterdam: John
Benjamins.
2002. Nonfinite Structures in Theory and Change. Oxford:
Oxford University Press.
Moussy, Claude. 1989. “Les métaphores lexicalisées et l’analyse
sémique.” In Lavency and Longrée (1989: 309-19).
Nielsen, Hans Frede and Lene Schøsler (eds.). 1996. The Origins and
Development of Emigrant Languages: Proceedings From the
Second Rasmus Rask Colloquium, Odense University, November
1994. (= NOWELE Supplement, Vol. 17). Odense: Odense University
Press.
Niemeier, Susanne. 2000. “Straight from the Heart ― Metonymic and
Metaphorical Explorations.” In Barcelona (2000a: 195-213).
Nyrop, Kristoffer. 1913. Grammaire historique de la langue française.
Vol. iv: Sémantique. Leipzig: Harrassowitz; Paris: Alphonse Picard;
New York: Stechert.
Partridge, Eric. 1983. Origins: A Short Etymological Dictionary of
Modern English. New York: Greenwich House.
Pons Sanz, Sara María. 2003. “Law and Literature: Norse-Derived
Vocabulary in the Works of Wulfstan, Archbishop of York and
Bishop of Worcester, With Particular Focus on the Legal Terms.”
Ph.D. thesis, University of Cambridge.
Rinelli, Gabriele. 2001. “Scandinavian and Native Social Terms in Middle
English: The Case of Cherl/Carl.” In Kastovsky and Mettinger
(2001: 267-77).
Rothwell, William. 1996. “Adding Insult to Injury: The English who
Curse in Borrowed French.” In Nielsen and Schøsler (1996: 41-54).
Rundbald, Gabriella, and David B. Kronenfeld. 2000. “Folk-Etymology:
Haphazard Perversion or Shrewd Analogy?” In Coleman and Kay
180 D. Gary Miller

(2000: 19-34).
Serbat, Guy. 1975. Les Dérivés nominaux latins à suffixe médiatif. Paris:
Les Belles Lettres.
Sihler, Andrew L. 1995. New Comparative Grammar of Greek and Latin.
Oxford: Oxford University Press.
2000. Language History: An Introduction. Amsterdam: John
Benjamins.
Shipley, Joseph T. 1984. The Origins of English Words: A Discursive
Dictionary of Indo-European Roots. Baltimore: John Hopkins.
Skoda, Françoise. 1988. “Les métaphores zoomorphiques dans le
vocalbulaire médical, en grec ancien.” Logopédies. 221-34.
Stern, Gustaf. 1931. Meaning and Change of Meaning: With Special
Reference to the English Language. Bloomington: Indiana University
Press. (Reprinted 1968).
Stockwell, Robert, and Donca Minkova. 2001. English Words: History
and Structure. Cambridge: Cambridge University Press.
Sweetser, Eve E. 1987. “Metaphorical Models of Thought and Speech: A
Comparison of Historical Directions and Metaphorical Mappings in
the Two domains.” Berkeley Linguistics Society 13: 446-59.
1990. From Etymology to Pragmatics: Metaphorical and
Cultural Aspects of Semantic Structure. Cambridge: Cambridge
University Press.
Tournier, Jean. 1985. Introduction descriptive á la lexicogénétique de
l’anglais contemporain. Paris: Champion-Slatkine.
Trask, Robert L. 1996. Historical Linguistics. London: Arnold.
Traugott, Elizabeth C. 1989. “On the Rise of Epistemic Meanings in
English: An Example of Subjectification in Semantic Change.”
Language 65: 31-55.
Uhlich, Jürgen. 2002. “Verbal Governing Compounds (Synthetics) in
Early Irish and Other Celtic Languages.” Transactions of the
Philological Society 100: 403-33.
 Semantic Change 181

Ullmann, Stephen. 1957. The Principles of Semantics. 2nd edn, 3rd

impression, with updated bibliography (1963). Oxford: Basil
Blackwell. (Repr. 1967.)
1962. Semantics: An Introduction to the Science of Meaning.
Oxford: Basil Blackwell.
Vries, Jan de. 1977. Altnordisches etymologisches Wörterbuch. 2nd edn.
Leiden: Brill.
Wackernagel, Jacob. 1926-28. Vorlesungen über Syntax, mit besonderer
Berücksichtingung von Griechisch, Lateinisch und Deutsch. 2 vols.
Basel: Birkhäuser.
Warren, Beatrice. 1992. “What Euphemisms Tell Us About the
Interpretation of Words.” Studia Linguistica 46: 128-72.
1998. “What is Metonymy?” In Hogg and van Bergen (1998:
301-10).
Watkins, Calvert. 1962. “Varia ii. 1. Irish Milchobur’. Ériu 19: 114-6.
(Repr. in Watkins 1994: i. 92-6 [#18]).
1994. Selected Writings, ed. Lisi Oliver. 2 vols. Innsbruck:
Institut für Sprachwissenschaft, University of Innsbruck.
(ed.) 2000. The American Heritage Dictionary of Indo-
European Roots. 2nd edn. Boston: Houghton Mifflin.
Williams, Joseph M. 1986 [1975]. Origins of the English Language. New
York: Free Press.
182 D. Gary Miller

ABSTRACT

Semantic Change

D. Gary Miller
(University of Florida)

The main issues in semantic change are examined. It is argued that

the various formulations and quasi-formalizations in the literature are
premature for the simple reason that semantic change is determined by
links in the human mind/brain. Mental schmata subsume an array of
affiliated conceptions that change in response to stimuli, as reflected in
meaning differences from speaker to speaker and over time.
In our attempts to classify and make sense of the universe, we place
value judgments on objective labels. These subjective (e)valuations differ
sometimes greatly from one speaker to another and as our conception of
the referents evolves over time. Part of (e)valuation involves
comparisons or utilization of a label from an item conceptualized as
similar. Metaphors at least initially bear evaluative nuance. Metonymy
involves transfer of labels by reanalysis mediated by associated
contexts. Attempting to itemize all forms of metonymy is no more
feasible than taxonomizing mental interconnections.
Whatever structure there is to semantic change is not even
constrained solely by the structure of links in the human mind/brain,
because changes in the environment and interaction with various
subcultures constitute independent variables that shape the outcome of
semantic change.

Key Words: semantic change, metaphor, metonymy, expressivity,

polysemic extension, melioration, pejoration