Also taken from rhetoric is the term irony, i.e. expression of one's meaning by words of opposite meaning, especially a simulated adoption of the opposite point of view for the purpose of ridicule. One of the meanings of the adjective nice is 'bad', 'unsatisfactory'; it is marked off as ironical and illustrated by the example: You've got us into a nice mess! The same may be said about the adjective pretty: A pretty mess you've made of it!
Changes depending on the social attitude to the object named, connected with social evaluation and emotional tone, are called amelioration and pejoration of meaning. Amelioration or elevation is a semantic shift undergone by words due to their referents coming up the social scale. For instance OE cwen 'a woman'> ModE queen, OE cniht 'a young servant' > ModE knight. The words steward and stewardess (the passengers' attendant on ships and airliners) have undergone a great amelioration. Steward < OE stigweard from stigo 'a sty' and weard 'a ward', dates back from the days when the chief wealth of the Saxon landowner was his pigs, of whom the stigweard had to take care. The meaning of some words has been elevated through associations with aristocratic life or town life. This is true about such adjectives as civil, chivalrous, urbane.
The reverse process is pejoration or degradation; it involves a lowering in social scale connected with the appearance of a derogatory and scornful emotive tone reflecting the disdain of the upper classes towards the lower ones. A knave < OE cnafa \\ Germ Knabe meant at first 'boy', then 'servant', and finally became a term of abuse and scorn. Another example of the same kind is blackguard. In the lord's retinue of Middle Ages served among others the guard of iron pots and other kitchen utensils black with soot. From the immoral features attributed to these servants by their masters comes the present scornful ' meaning of the word blackguard. A similar history is traced for the words boor, churl, clown, villain.
Euphemism (Gr euphemismos from eu 'well' and pheme 'speak') is the substitution of words of mild or vague connotations for expressions rough, unpleasant or for some other reasons unmentionable.
Within the diachronic approach the phenomenon has been repeatedly classed by many linguists as taboo. This standpoint is hardly acceptable for modern European languages. With primitive peoples taboo is a prohibition meant as a safeguard against supernatural forces. Names of ritual objects or animals were taboo because the name was regarded as the equivalent of what was named. S. Ullmann returns to the conception - of taboo several times illustrating it with propitiatory names given in the early periods of language development to such objects of superstitious fear as the bear (whose name originally meant 'brown') and the weasel. He treats both examples as material of comparative semantics. The taboo influence behind the circumlocutions used to name these animals becomes quite obvious when the same phenomenon is observed in similar names in various other languages. There is no necessity to cite them here as they are given in any book on general linguistics. It should be borne in mind that taboo has historical relevance. No such opposition as that between a direct and a propitiatory name for an animal, no matter how dangerous, can be found in present-day English.
With peoples of developed culture, euphemism is intrinsically different, has nothing to do with taboo and is dictated by social usage, moral tact and etiquette. Cf. queer 'mad', deceased 'dead', perspire v 'sweat'.
From the semantical point of view euphemism is important because meanings with unpleasant connotations appear in words formerly neutral, as a result of their repeated use instead of other words that are for some reason unmentionable.
The material of this chapter shows that semantic changes are not arbitrary. They proceed in accordance with the logical and psychological laws of thought, otherwise changed words would never be understood and could not serve the purpose of communication. The various attempts at classification undertaken by traditional linguistics, although inconsistent ( and often subjective, are useful, since they permit the linguist to find his way about an immense accumulation of semantic facts. However, they say nothing or almost nothing about the causes of these changes.
CHAPTER II. CAUSES OF SEMANTIC CHANGE
In comparison with classifications of semantic change the problem of their causes appears neglected. Opinions on this point are scattered through a great number of linguistic works and have apparently never -been collected into anything complete. And yet a thorough understanding of the phenomena involved .in semantic change is impossible unless the whys and wherefores become known. This is of primary importance as it may lead eventually to a clearer, interpretation of language development. The vocabulary is the most flexible part of the language and it is precisely its semantic aspect that responds most readily to every change in the human activity in whatever sphere it may happen to take place.
The causes of semantic changes may be grouped under two main headings, linguistic and extralinguistic ones. Of these the first group has suffered much greater neglect in the past and it is not surprising therefore that far less is known of it than of the second. It deals with changes due to the constant interdependence of vocabulary units in language and speech, such as differentiation between synonyms, changes taking place in connection with ellipsis and with fixed contexts, changes resulting from ambiguity in certain contexts, and some other cases.
Semantic change due to the differentiation of synonyms is a gradual change observed in the course of language history, sometimes, but not necessarily, involving the semantic assimilation of loan words. Consider, for example, the words time and tide. They used to be synonyms. Then tide took on its more limited application to the periodically shifting waters, and time alone is used in the general sense.
Another example of semantic change involving synonymic differentiation is the word twist. In OE it was a noun, meaning 'a rope' whereas the verb thrawan (now throw) meant both 'hurl' and 'twist'. Since the appearance in the Middle English of the verb twisten ('twist') the first verb lost this meaning. But threw in its turn influenced the development of casten (cast), a Scandinavian borrowing. Its primary meaning 'hurl', 'throw' is now present only in some set expressions. Cast keeps its old meaning in such phrases as cast a glance, cast lots, cast smth. in one's teeth. Twist has very many meanings, the latest being 'to dance the twist'
Fixed context may be regarded as another linguistic factor in semantic change. Both factors are at work in the case of token. When brought into competition with the loan word sign, it became restricted in use to a number of set expressions such as love token, token of respect and so became specialized in meaning. Fixed context has this influence not only in phrases but in compound words as well. OE mete meant 'food', its descendant meat refers only to flesh food except in the set expression meat and drink and the compound sweetmeats.
No systematic treatment has so far been offered for the syntagmatic semantic changes depending on the context. But such cases do exist showing that investigation of the problem is important.
One of these is ellipsis. The qualifying words of a frequent phrase may be omitted: sale comes to be used for cut-price sale, propose for to propose marriage, to be expecting for to be expecting a baby. Or vice versa, the kernel word of the phrase may seem redundant: minerals for mineral waters. Due to ellipsis starve which originally meant 'die' (cf. Germ sterben) came to substitute the whole phrase die of hunger, and also began to mean 'suffer from lack of food' and even in colloquial use 'to feel hungry'. Moreover as there are many words with transitive and intransitive variants naming cause and result, starve came to mean 'to cause to perish with hunger'.
English has a great variety of these regular coincidences of different aspects, alongside with cause and result, we could consider the coincidence of subjective and objective, active and passive aspects especially frequent in adjectives. E.g. hateful means 'exciting hatred' and 'full of hatred'; curious'strange' and 'inquisitive'; pitiful 'exciting compassion' and 'compassionate'. Compare the different use of the words doubtful and healthy in the following: to be doubtful :: a doubtful advantage, to be healthy :: a healthy climate.
The extralinguistic causes are determined by the social nature of the lang