11. Sentences (2) In the previous chapter we characterised sentences.as_combinations of a verb and one or more nouns, which may in turn have associated with iherrTmodifiers~(ě g adjectives) and be linked to the verb bj\a relational device>(preposition or case" inflecti&a)_Sentences were distinguished byfype. according to the number of nouns determined by the verb and their relationships to the verb and in some cases to each other. We will now identify ancLgive labels to the functions that the verb and the nouns perform in the structure of sentences:'---------""'' ^--—............_____- Functions ________ The function performed-by-the verb in a sentence we refer to as the AEtedjcatoť function: the verb 'predicates' an action, process or state of a noun or nouns. The element that we referred to as the 'first noun' in the previous chapter performs the 'Sub-jeet^function in a sentence: it is the agent of an action Predicator, the under-goer-afja process,-the :'thing' of which a location or state is predicated. There is often person and number agreement (see Chapter 16) between the noun functioning as Subject and the verb functioning as Predicator. Consider the present tense paradigm of the verb travailler 'work' in French: je_trayaille, tu travailles, il/eHe^travaille, noun travaillons, vousjravaillez, ils/elles travaillent. That is to say, "tu travailles" is analysed as 'Subject:noun—Predjeator:verb', with the categories of '2nd person'/'singular number' of the Subjecf tu marked in the Predicator travailles by means of the -es suffix—working here Wfn the written rather than the spoken forms. X, ■ The 'second noun' in a sentence may perform a number of functions, depending on the sentence type in which it occurs. In a transitive sentence the second noun performs the function of^Cibject—in the sentence: it represents the undergoer of the action, the 'object' towards which the action is directed, eg the camp-firejn "The boy^c^uJJirtiie-.camp-rire''. The second noun in a stative or equative sentence has the syntactic function of 'Complement',- eg...in sentences like "The journey was a frightening experience", "Harry is the carpenter". The Complement referTbacFto and elaborates or "describes the Subject of the sentence. As we have seen, in stative sentences the Complement function may be performed by an adjective rather than a noun, eg "The journey was frightening"'. The second nounin sentences of the locative or temporal type functions as an 'Adjunct', representing additional, circumstantial information about a Subject or a predicated action or event, as in r'The children ire in the garden", "Our guests departed the day before yesterday". Second nouns{ then, may function "syntactically as Object, Complement or Adjünct,-aecording to the sentence type, or ultimately according to the meaning of the verb functiöjiing-as^Predicatox_._ We turn now to the syntactic functions of a possible third noun in a sentence. In the 59 60 Discovering grammar previous chapter we mentioned one sentence type that contained a third noun: the ditransitive type, eg "We sent all our friends a postcarrJIl/^We sent a'posteard to all our friends", where (to) all our /ní^Fts-ccrfršíHered-rhe^third-nounT that the third noun usdíuTyTTäTttie~semantic function of^recipienť or 'benef Syntactically, the third noun in ditransitive sentences functioTrs^avOtrfěct; that is to say, in di|rajisitLye„s.en.tences there are two nouns functioning as Object. A distinction is sometimes drawn between the 'Direct Object' (the second nouft^éŕrécth/ involved in the action) and the 'Indirect Object' (the third noun, only 'indirectly' involved)-^" There are two further not uncommon sentences-types that contain a third noun. The first can be illustrated fromEnglish,by_iJieToUowing sentence: "We thougfit_Harry's^ deaih-3 terrible tragedy". Here we is analysed as Subject, thought as"Predicator, Harry's death aTObJect, and a terrible tragedy as Ckmiplemejtt7.TTh&^Coniplemeht now refers backjtp and elaborates the Objeet (rather than the Subject asTimative sentences). This sentence type could perhaps be called a 'transitive/st^tiyďtypé. The^seeond further type with a third noun can be illustrated from English by: 'The carpenter keeps his tools in a leather bag". Here the carpenter is analysed as Subject, keeps as Predicator, his tools as Object, and in a leather bag as- Adjunct. The Adjunct in this sentence type gives information on the location of the Object (rather than of the Subject as in the locative type). This type coulď perhaps be caIled_A^trAnsitive/locaIü/el^ex^^ Both sentence types just discussed are sometimes referred to as 'complex-transitive'./ Identifying syntactic functions--------""" We have so far attempted to characterise syntactic functions by alluding to their role in the economy of particular sentence types. Another identification procedure is to examine the kinds of questions that relate to each of theTunciions. The Subject of a sentence is elicited by questions like: "Who/Whfififfdid something?", "What happened?" The Object of a sentence is elicited by: "What/Who{m) did someone Ůffror "To/For Who{m) did someone do something?" (Indirect Object). The " Adjunct of a sentence is elicited by: "Where/ When/How [Why did someone do something?" or ". .. did something happen?" or "... is something?". The Complement of a sentence is elicited by: " What is someone/something?" This procedure may be combined with one that starts with the verb functioning as Predicator and asks what additional elements are determined by that particular verb (cf Chapter 10). For example, the verb send could be said to determine someone who sends {"Who jsends?"), something that is sent {"What did someone send?"), and someone to whom it is sent ("To whom did someone send something?") OTjotmEBESlil jsjient(-"W7iere did sSmeone sehcTšomething?"). That is, send enters either a ditransitive sentencetype^contäinmg two Objects) or a transitive/locative sentence type (containing Object and Adjunct). \ Funciions-an4*enťeííce types It is now possible to look again at sentence types and to describe themjnjerms of the-syfttacť^_functionaUl0te-that each of them opens up. The intransitive sentence Sentences (2) 61 type, for example, opens up a Subject and a Predicator slot only, the ditransitive type a Subject, a Predicator and two ÔlíječTslotSv.We can summarise the structure 0f sentenee-typ'esThaTwe have discussed in the following table (where S is Subject, p is Predicator, O is Object, C is Complement, A is Adjunct):-------- Intransitive /ST Stative or Equative / S P C \ / (L-^e Locative or Temporal SPA \ Transitive S P O \ Ditransitive j S P O O Transitive/Stative S P O C i Transitive/Locative \ S P O A / It should be noted that the list is not complete,,iii the sense that many languages have more, fewer, or differerjtsentencejypeS''than these. Also, the order of the elements in the sentence types abovVcorresponds to the Lneutraľ_7order for English, and other languages that place Subject before Predicatór'änď/'ObjectSxó^CompIe-ment after the Predicator. Other languages have a differentxíoeutraľ or 'basic' order; eg Western Desert (Australia) has SOP order: "watilu maku kultunu" (lit. 'man kangaroo speared').-----------------" Exercise 24 Identify for the following English sentences which sentence type is represented, and indicate which words belong to each functional position; eg "Bill (S) sent (P) his wife (O) a message (O)"—ditransitive. ------------------------' ~~""~^~—________ 1. The shipyard is building a new oil-tanker. 2. Harry is sitting in the garden. 3. The children will put their muddy boots on the kitchen floor. 4. Susan is a first-class journalist. 5. Last night's storm blew over the tree in the corner. 6. The committee has appointed Edward as its secretary. 7. Our parking time expired five minutes ago. 8. Harry was telling us a funny story. 9. The branch is breaking. 10. My coat is the brown one-Obligatory and optional elements We have discussed the structure so far in terms of a verb and a number-of nouns, functioning in positions opened up according to sentence type, or alternatively, determined by the 'meaning* pf the verb. The implicationJias-been-lhaLi^yerb 'requires' a specific number of nouns in order for the sentence_to_be 'grammatical' or 'complete', or to 'make sense'. We havenuted, eg in the case of English send, that a verb may function as, Predicator in more than one sentence type (ditransitive^ and transitive/locative for send). But-alLthe-iuncti&nalr-positions determined by the verb or implied by the sentencejtyple have.been filled_._That is to say, we have been dealing so far only with /obligatory/elements of sentences. I......- - ý There are two senses in which an element of sentence structure may be regarded as 'optional' rather than 'obligatory'. First of all, it is possible to add elements to / 62 Discovering grammar sentences that are not directly required by the verb. Consider: "The agent gave me the parcel yesterday in the park". Give'is a ditransitive verb; it requIfeT^Subjeet (the agent) and two Objects (me, thej^rcel). The sentence would be grammatical or complete with just these slots filled: yesterday and in the park are thus gratuitous elements that have been added to the basic sentence type. Note that they are both functioning as Adjunct, as a temporal and a locative Adjunct respectively. It is possible to specify time and place for almost any event, but it is arguable that few verbs actually require them to be specified. Optional Adjuncts—of time, place, manner, reason, purpose, etc—may be freely added to almost any sentence, depending on the contraints of context. The second sense in which it is possible to speak of an 'optional' element in sentence structure entails the omission of an element that is regarded as being required by the verb. Consider the verb write in English. Arguably this verb enters a ditransitive sentence type, ie 'someone write something to/for someone' (eg "Harry is writing a letter to his aunt", "Harry is writing a report for the director"). But it is possible to omit either or both of the Objects,"eg^Harry is writing" (in answer to the question "What's Harry doing?"), "Harry is writing a letter", "Harry is writing to his aunt". In the case of a verb like write we must say either that write enters the intransitivejyidLiransittve sentence types as well as the ditransitive, or thsťwrife is basicallý'ditransitrvejbut that either or ^ Clearly, this is a different kind of optionality from the firstjtmdfand'ft may"be more appropriate to distinguish it by referring to it as 'deletabilijV' rather than optionality. That is, the Objects of write are said to be ^deletable' according to the conditions of context; for it is usually the case that context determines or allows the dejetabjlity of otherwise obligatory elements of sentence structure. It is probably also the case that a semantically restricted Object is more likely to be deleted than one not so.restricted; for example, Indirect Objects nearly always refěTTíPpeísons and are thus probably more often deletable, and in the case of a verbTike write-, the Direct Object is more-or-less restricted to a set of nouns referring to^written artefacts (cf also read, sing, play). Functions and classes ^We began our discussion of sentences^ by as^erťhig_that a sentence is essentially made up-of a verb an.d-soflie accompanying nouns. We havěnhaďtb revise that /assertion in th^intervening page«, in order tpaccôunt for verbless sentences and to aHow for adjectives appearing instead of nouns. Having now irítfbdučědTříě functional slots of sentence structure, we need to bring the functions into relationship with the classes (or categories) of word (ie noun, verb, etc). The Predicator functional slot is filled by a verb, along with any of its accompanying modifying elements (eg auxiliary verbs, negative particles), ie by what might be called a 'verb phrase'. The Subject slot is usually filled by a noun, along with its accompanying modifying elements (see Chapters 12 and.13), ie by ar'noun phrase', or by a pronoun or proper noun (name). Pronouns and proper nouns are not usually accompanied by any modifying elements, and in any case the possibilities are normally restricted. For example, in English a pronoun may be postmodified by Sentences (2) 63 a relative clause (see Chapter 13): "he who hesitates". But compare the equivalents of this in French and German: "celui qui hésite" 'the one who hesitates', ie not using the pronoun U 'he'; "wer zögert'"who hesitates', ie not using the.pronoun er 'he'. The Object slot is also usually filled by a noun phrase or pronoun or proper noun. It may additionally be accompanied by a preposition or postposition, eg "I gave it to John", French "Je ľai donne ä Jean"; Hindi "us se yah saval puchie", literally 'hirn-to this que_stion ask' (ie "Ask him this question"). The Complement slot is usually filled either by a'noun phrase or by an adjective and any accompanying modifiers it might have (see Chapter.. 14), ie by an 'adjective phrase*.These may also be accompanied by a preposition or postposition, eg "I regard hirr^ as a friend". Finally, the Adjunct slot may be filled by a number of categories: a noun phraseLanoun phrase accompanied by a preposiťiompť postposition, or an adverb; eg Hindi "yah apne ghar laut gaya" 'he his home back went' (ie "he returned to his home"), "budhvar ko ao" 'Wednesday on come' (ie "Come on Wednesday"), "vah acchagati hai" 'she well singing is' (ie "she sings well"). ._____. \ It is thus possible to describe the structure of a sentence by identifying_bath-the y functional slots and the categories of word or phrase that fill them. For example: "The keeper (S: Noun Phrase) is giving (P: Verb Phrase) the lions (Oi: Noun Phrase) their meat (Od: Noun Phrase)". "Our friends (S: NP) live (P: VP) in Brussels (A: Prepositional Phrase)". "This room (S: NP) feels (P: VP) very cold (C: Adjective Phrase)". j ("J j , /jUt^r Exercise 25 -^UX fr^'J-1- For the following data (SIL 1980: E13) from Tlingit (Alaska), make an analysis""oF each sentence identifying functional slots and categories of phrase that fill them, and indicate which sentence type each belongs to. 1. xóots saxwaa.áx dzeeyáak 'I heard a brown bear earlier on' brown-bear voice-I-heard earlier-on 2. Juneau-dé kukgwaatěen 'He's going to Juneau on a trip' he-will-take-a-trip 3. kúnax kusi.áať 'It's really cold' really it-is-cold 4. tlax a yáanax ee wdixwétl 'You're too tired' too-much it face-beyond you is-tired 5. ee xöonee • 'It's your friend' your friend 6. aatlěindáanaadoojěewoo 'He has lots of money' lots money his hand-locative marker 12. Expandingthenound) Noun phrase We referred in the previous chapter to the possibility of a noun, functioning at a particular position in sentence structure, being accompanied by one or more modifying elements. And we termed this combination a 'noun phrase'. Some descriptive linguists would recognise a phrase level of syntactic structure intermediate between word and sentence. That is, the structure of a sentence is described in terms of its constituent phrases, and these phrases in turn are described in terms of their constituent words. So the general definition of a 'phrase' becomes: "a unit consisting of one or more words". In such a descriptive framework each unit is described in terms of the units at the level immediately below. In this book we have taken the view so far that sentences are considered to be combinations of words, which may in turn be accompanied by modifying items. That is to say, we are regarding 'sentence' and 'word' as the primitive terms in our syntactic description. And, as the titles of this chapter and the next two imply, we are regarding phrases as expansions of particular classes of words. The class of nouns is the one, probably in all languages, that may be subject to the most and to the most varied expansion by modifying elements. In this chapter we consider the expansion of nouns by means of members of other word classes, and in the next chapter we look at modifiers that are themselves syntactic structures. Closed class modifiers Towards the end of Chapter 2 we identified a number of word classes that we termed 'closed' rather than 'open', because their membership is restricted in number, changes only very slowly over time, and can easily be listed exhaustively. Some of these classes function in the expansion of nouns, especially the class of Determiners. The membership of the Determiner class varies from language to language, and it includes a number of quite diverse subsets of items, all of which, however, contribute to 'determining' the contextual status of a noun. First of all, a distinction is often made in the Determiner class between 'Identifiers' and 'Quantifiers'. As the latter label implies, these items have the function of specifying 'how many' or 'how much' of a particular noun is being referred to. Such a specification may either involve an actual number ("five coaches", "the third coach") or be an expression of indefinite quantity ("several coaches", "some cheese"). Clearly, there is a relation of compatibility here between quantifiers and the subclasses of 'mass' and 'countable' nouns. For example, some is compatible with the singular of mass nouns ("some cheese"), but with the plural of countable nouns ("some coaches"); and 64 Expanding the noun (1) 65 numerals are compatible with countable nouns only, unless they precede a specific expression of quantity ("five pounds of cheese", "twelve litres of petrol"). Compare German "einige Wagen" 'some coaches', "etwas Käse" 'some cheese'; and French "des wagons" 'some coaches', "du fromage" 'some cheese'. The class of Identifiers includes, in English, the 'articles' (indefinite a, definite the), the demonstratives (this, that), the possessives (my, your, his, etc). In English these items are mutually exclusive, ie the occurrence of one excludes the possibility of the occurrence of the others in the same noun phrase; eg *"my the box" is not a possible noun phrase in English. Many languages (eg Latin) do not have items equivalent to the English articles: 'definiteness' is signalled in other ways. Neither do all languages show the same mutual occurrence restrictions as English. Consider the following data (SIL 1980: B2) from Bekwarra (Nigeria): 1. ugam '(the/a) mať 2. 3. ugam iyi ugam it Jia 'my mať 'three mats' 4. 5. ugam abin ugam iyi itj*ia abin 'this mať 'these three mats of mine' Note that the modifiers of the noun come after it in Bekwarra, whereas the equivalent items in English come before the noun, except for of mine in the last example. This, however, is the English way of including a demonstrative and a possessive in the same noun phrase (cf *"my these mats"), a structure that is possible with the regular possessive identifier in Bekwarra. Note also, incidentally, that Bekwarra does not mark 'plural' number in either the noun or the demonstrative, as English does. A further descriptive point is the relative order of the modifiers in the noun phrase. From Item 5 we can describe the structure of the noun phrase in this Bekwarra data as: noun—possessive—numeral— demonstrative. Exercise 26 Describe the structure of the noun phrase in the following Agatu (Nigeria) data (based on SIL 1980: B5): 1. ugwuoye 'one hen' 2. ugwueho 'five hens' 3. ugwudu 'all the hens' 4. ugwuehodub 'all those five hens' Open class modifiers The most obvious and extensive class in this group is that of Adjectives, perhaps the set of words that most readily springs to mind in the context of noun modification. Adjectives referring to colour, size, shape, texture, provenance, etc, are widespread in the languages of the world. Very often they can function both as modifiers of nouns and in the predicate of a clause as Complement, eg "the red hat", "the hat is red". As we have noted already (Chapter 10), descriptive or Stative sentences like this are frequently verbless. Within the noun phrase, some Ian- 66 Discovering grammar guages (like English) place the adjective before the noun, while others place it after, and yet others have adjectives in both positions. A language of this last kind is French, eg "le chapeau rouge" 'the hat red', "le bon enfant" 'the good child'; although noun-adjective is the more frequently occurring order, with the alterna' tive limited to a relatively small set of common adjectives. English also has an adjective-following pattern, though not as a rule with nouns: indefinite pronouns, however, are normally followed by an adjective modifier, eg "somebody neutral", "anything unusual", "nothing alcoholic"; but cf "a little something", "an absolute nobody"—where the pronoun has more or less taken on the status of a noun. When more than one adjective occurs as modifier of a noun, it may be the case that classes of adjective are ordered relative to each other. In English, for example, we would be more likely to say "the big red hat" than ?"the red big hat"—unless there were two 'big hats' and we wished to distinguish the red one from one of another colour, in which case red would carry contrastive stress. In English, then, adjectives referring to size normally precede those referring to colour. Compare French "le grand chapeau rouge", where the size adjective grand 'big' precedes the noun and the colour one, rouge 'red', follows it. It would appear that, in English, the more criterial the adjective is in distinguishing the noun, the closer it is put to it in the noun phrase, so that colour is normally considered more criterial than size. Consider now the following data (based on SIL 1980: B7) from Mambila (Nigeria): 1- těl 'elephant' 2. tělbúnu 'this elephant' 3. tělachi 'your elephant' 4. tělamohbúnu 'this elephant of mine' 5. telamöhdua 'my large elephant' 6. tělduayili 'a large black elephant' 7. těl bonu ai 'these two elephants' 8. těl bonu duayilifäl 'these two large black elephants' From this data we can conclude that the structure of the noun phrase in Mambila may be expressed, at least partially, by the following formula: noun—possessive—demonstrative—adjective (size)—adjective (colour)—numeral. Here, too, the size adjective precedes the colour one, but is closer to the noun; though in Mambila the adjectives are separated from the noun by closed-class modifiers. We have been talking of adjectives as if they were a universal class of words. This is by no means the case. Some languages have a class of modifying words that corresponds to the classes of adjective and adverb in English: compare the Apinaye (Brazil) data in Exercise 5 (Chapter 2). This is also the case, more or less, in German, where it can be said at least that there is a large overlap between the set of open-class modifiers of nouns and the set of items functioning as Adjunct in the predicate structure of sentences; cf "die schöne Sängerin" 'the beautiful singer', "sie singt sehr schön" 'she sings very beautifully'. What is widespread, if not indeed universal, is the occurrence of a set of open-class words that have as one of their functions the modification of nouns. The other set of open-class modifiers of nouns is perhaps much less widespread: it i Expanding the noun (1) 67 nouns themselves, functioning as modifiers of other nouns. For example, English "the felt hat", where the noun felt modifies hat. Arguably, this noun + noun combination (felt hat) could be considered as a compound (see Chapter 7), but it does not have the characteristic single primary stress on the first element (cf fire-bucket). Moreover, the modifying noun often co-varies with adjectives having similar reference (cf "the woollen hat") and the combination is not felt semantically to be a single unit. Possibly, however, there is considerable variation in degrees of cohesiveness between noun + noun compounds at one end of the«cale and noun-modifier + noun combinations at the other. Defining/non-defining It is probably true to say that adjectives more often than not have a defining role; that is, they serve to assign the noun being modified to one subset of such nouns rather than another, or they distinguish the 'thing' being referred to by the modified noun from another possible 'thing' referred to by the same noun. For example, the adjective red in "the red hat" assigns this hat to the subset of hats distinguished by their being red; in "my red hat", red distinguishes this hat from hats of other colours that I might have. Such uses of adjectives are said to be 'defining'. If, however, I have only one hat, then to refer to it as "my red hat" is to provide gratuitous and non-essential information. Indeed the adjective red would normally be understood as having a defining function in the phrase "my red hat"; and one would probably resort to other strategies if one intended red as non-defining, eg "I can't find my hat. It's red", where the indication of colour is intended as a clue to establishing its whereabouts. Arguably, in the expression "It's red", the adjective red, although no longer a modifier, has a defining function, assigning "my hat" to the subset of hats coloured red. In the phrase "my lovely wife", however, the adjective lovely would normally be taken to have a non-defining function, assuming that it is spoken in a monogamous society! Similarly, the adjective late is non-defining in the phrase "his late father"; and the adjective old in the phrase "the old railway worker", where old is added merely as a characterising feature of the noun, not as a defining one. Note in this example, though, that the noun-modifier railway does have a defining function, assigning the noun worker to a particular subset of workers. The distinction between defining and non-defining will again be important in the discussion of clausal modifiers, especially relative clauses. Definite vs indefinite Many languages, including English, have overt means of marking a noun phrase as 'definite' or 'indefinite'. This is often done by means of a set of Determiners called 'Articles': a and the in English, 'indefinite' and 'definite' article respectively. The primary function of the articles is in the relationship between the sentences of a text or discourse. The indefinite article marks a noun as 'first mention' of a referent, while the definite article marks the noun as 'already introduced' into the discourse. For example, in the sentence "A Japanese team has climbed the mountain", the 68 Discovering grammar noun mountain is marked as a referent already under discussion, while Japanese team here receives its first mention. In languages which do not have items equivalent to the articles of English (eg Russian, Bimoba), other means are used to indicate that a noun is to be interpreted as definite or indefinite. In Russian, for example, the positioning of a noun in the initial slot of a sentence marks it as definite. In Bimoba, special particles occur after nouns that have been mentioned before in the discourse. And in Lithuanian there is a special pronominal form of adjectives, used with nouns that have already been introduced into the discourse. Another function that the definite article has (in English, for example), is to occur with nouns that are modified by a defining relative clause or prepositional phrase; eg "the woman that I saw you with yesterday", "the house with the red door". Note that both "a house with the red door" and "the house with a red door" sound odd in English, but not "a house with a red door". Exercise 27 For the following data (SIL 1980: Bl) based on Coatlán Mixe (Mexico) describe the structure of the noun phrase: 1. po:f tsu:tf ?ifp 'The spider sees the horsefly' spider horsefly sees 2. he po:f tsu:t| ka:jip 'That spider will eat the horsefly' that spider horsefly will-eat 3. tsu:k t/i:t ?ifti The rat saw the cat' rat cat saw 4. tfi:tpo:p tsu:k?ijp 'The cat sees the white rat' cat white rat sees 5. tJ"i:tpo:p tsu:kjah?o\>kip 'The cat will kill the white rat1 cat white rat will-kill 6. po:p tJLthe po:p tsu:kka:jti 'The white cat ate that white rat' white cat that white rat ate 7. metstfi:tmetspo:p tsu:kka:jti 'Twocatsate two white rats' two cat two white rat ate 8. mahpo:/ tsu:kjah?o?okti 'The big spider killed the rat' big spider rat killed 9. mets man tji: t mets mah tsu: k jah?o?okti 'Two big cats killed two big rats' two big cat two big rat killed 10. he mets tsu: t he tsu:k^ijti 'These two horseflies saw that rat' this two horsefly that rat saw 13. Expanding the noun (2) In the previous chapter we considered ways of expanding the noun with single word items: determiners and adjectives. In this chapter we turn our attention to the expansion of nouns by means of structures comprising more than one word: phrases and clauses. Possessive phrase possessive phrases allow a relationship of possession to be expressed between one noun (phrase) and another noun (phrase); eg English "the big man's overcoat". The possessive phrase in English takes up the position of the identifier in the modification of the noun; more specifically, it replaces a possessive identifier, cf "his overcoat", "the big man's grey overcoat". It will be noted that the possessive noun phrase in English is marked by the possessive (genitive) case suffix 's. In fact, this is not a genuine suffix on the noun, since, as we noted in Chapter 9, it is attached to the last item of the possessive phrase: "the man in the corner's overcoat", "the last man out's overcoat". It is, therefore, more appropriately termed a possessive clitic. English has an alternative means of expressing the possessive relationship: by a noun phrase introduced by the preposition of, placed after the noun which is possessed, eg "the tail of the aeroplane" (cf "the aeroplane's tail"). The o/-possessive construction seems to be preferred with aeroplane, whereas the 's-clitic construction seems preferred with man, cf "the grey overcoat of the big man". In general, nouns referring to persons prefer the premodifying position in English, unless the noun being possessed is itself heavily modified, eg "the grey overcoat of the big man with the bowler hat talking to the lady in the fur-coat". Contextual factors also play a part in the choice of possessive construction in English. Few languages have the luxury of a choice of possessive phrase. In Punjabi, for instance, there is one way of forming a possessive phrase, which is a kind of synthesis of the two English means. Punjabi has postpositions instead of the prepositions of English: the possessive phrase is formed by means of the postposition da after the possessor noun (phrase). The possessive phrase (ie noun phrase da) is positioned in the premodifying position of the noun being possessed, eg "mwnde di kytab" (lit. 'boy of book'), 'the boy's book'. Consider now the followeing data (SIL 1980: G6) from Bariba (Dahomey): 1. sabii 'Sabi' 2. sabiinkuro 'Sabi's wife 3. duro 'man' 4. duro wi 'that man' 69 70 Discovering grammar 5. duroboko 'the big man' 6. durowinkuro 'that man's wife' 7. duro bokon kuro 'the big man's wife' 8. durogeowi 'that good man' 9. durogeowinkuro 'that good man's wife' 10. sabiinkurogeo wi 'that good wife of Sabi' From Nos 4, 5 and 8 it will be noted that determiners and adjectives normally follow (postmodify) the noun in Bariba. The possessive phrase in Nos 2, 6, 7, 9 and 10, however, precedes (premodifies) the noun, which means that it does not replace one of the other modifiers. Thus in No 10, English cannot have both that (demonstrative identifier) and Sabi's (possessive phrase) as premodiners, and has to resort to the circumlocution with the postmodifying o/-phrase. The marker of the possessive phrase in Bariba is a clitic -n, attached to the last item of the phrase: to the noun in Nos 2 and 10, to the demonstrative in Nos 6 and 9, and to the adjective in No 7. Exercise 28 Describe the expression of possession in the following Basari (Ghana) data (SIL 1980: G9) 1. uboti 'chief 2. uninjabotiu 'man's chief 3. unimpu 'wife' 4. uninjanimpuu 'a man's wife' 5, unimpu ubn 'one wife' 6. uninja-nee nimpuu ubo 'this man's one wife' 7. kusaau 'farm* 8. kusaau kubo 'one farm' 9. uninjasaaku 'a man's farm' 10. uninjasaaku kubo 'a man's one farm' 11. uninjaubosaaku 'one man's farm' 12. kukuntuu 'mortar' 13. ukuntuuku 'his mortar' 14. u nimpuu kuntuuku 'his wife's mortar' 15. uninja-nee nimpuu kuntuuku 'this man's wife's mortať Relative clause The relative clause is a means of expanding or modifying a noun by means of a whole sentence (or predication), without the noun itself having a function in the sentence, except through a substitute. Compare in English: "the girl is eating her breakfast", "the girl who is eating her breakfast". The first of these examples is a sentence, and the noun girl has a function in the sentence, viz that of Subject. The second example is not a sentence (it is a noun phrase), though it contains a sentence, viz "who is eating her breakfast"; and the noun girl has no function in this sentence except through the substitute (pronoun) who. Here, then, the sentence—termed a 'clause' because it is 'subordinate' or 'embedded'—is part of the noun phrase; it modifies and expands the noun. In English the relative clause occurs in postmodification position in the noun phrase. It is linked explicitly to the head noun by means of the noun substitute, the Expanding the noun (2) 71 relative pronoun, which always occurs in initial position in the relative clause, irrespective of its function in the clause. In the example above, the relative pronoun who is functioning as Subject in the relative clause, and so it occurs in its normal position in sentence structure. When relative pronouns occur with functions other than Subject, the normal structure of the sentence may be disturbed; eg in "the girl that I like best" the relative pronoun that functions as Object, so that the order in the clause is 'Object—Subject—Predicator . . .'. In fact, when a relative pronoun functions as Object in the relative clause in English, it may be omitted ("the girl I like best"). Note also that English has a relative pronoun that functions as a possessive phrase: whose, eg "the philosopher whose ideas I am quoting". Here, whose ideas is Object in the relative clause, and whose substitutes for the philosopher's as possessive phrase modifying ideas. The distinction that we discussed in the previous chapter between 'defining' and 'non-defining' is relevant also to relative clauses. A relative clause may be used to define the reference of a noun; eg "Which girl were you talking to?" "I was talking to the girl who was wearing the red scarf." Alternatively, a relative clause may be non-defining, providing additional, gratuitous information about the referent of a noun; eg "I was talking to a girl from the team yesterday .... The girl, who was wearing a scarf in the team colours, was telling me . . .." Non-defining relative clauses in English are usually bounded by commas in writing and are intonationally distinct in speech. The same convention does not apply in written German, though, where all relative clauses are bounded by commas: "die Dame, die einen roten Halstuch trug, hat mir gesagt ...", 'the lady who was wearing a red scarf told me .. .' or 'the lady, who was wearing a red scarf, told me In Hindi, a relative clause, together with the noun that it is modifying, often precedes the sentence in which the noun phrase has a function (as Subject, Object, etc). The noun modified by the relative clause is then 'picked up' in the main sentence by a 'correlative' pronoun, eg 'jis admi ne yah patr likha, vah bhartiy hoga" who man this letter wrote he Indian probably-is, ie 'The man who wrote this letter is probably an Indian'. In fact, the rule seems to be that definite nouns modified by relative clauses are placed before the main sentence, while indefinite nouns modified by relative clauses are placed after the main sentence, cf "maim ek admi se bat kar rahatha jo kal bharatjaega I a man to talk cont. was who tomorrow India will-go, le 'I was talking to a man who is going to India tomorrow'. The so-called 'relative-correlative' construction in Hindi includes a greater variety of structures than just relative clauses. 72 Discovering grammar Consider now the following data (SIL 1980: Gl) from Konkomba (Ghana): 1. u ba kpo 'He died' he past die 2. bi ba kan uwon 'They saw the rabbit' they past see rabbit 3. bi ba kan uwon u n ba kpo na 'They saw the rabbit which died' they past see rabbit it past die 4. unambuun ba gcen 'The cat slept' cat past sleep 5. u ba fii 'He/It got up' he/it past get-up 6. unambuun u n ba geen na ba fii The cat which slept got up' cat it past sleep past get-up The relative clause is placed immediately after the noun that it modifies. The modified noun is represented in the relative clause by the 3rd person singular pronoun h; in both examples, this functions as Subject in the relative clause. The relative clause is marked by the particles n and na: n occurs after the Subject pronoun, and na occurs clause-finally. Non-finite clauses One of the functions of non-finite clauses (ie clauses containing a non-finite verb form—infinitive or participle) in some languages is to modify nouns; eg in English ■'the girl looking at the book", "the accident witnessed by the people at the bus stop", "the person to see about your problem". In English, non-finite clauses postmodify the noun. In German, where they occur less frequently, they generally premodify the noun if they are participle clauses; eg "der den Ball mit der Hand berührende Fussballspieler" (lit. 'the the ball with the hand touching footballer'), ie 'the footballer touching the ball with his hand'; "die an der Haltestelle angehaltene Strassenbahn" (lit. 'the at the stop stopped tram'), ie 'the tram stopped at the stop'. Note that the participles come finally in the non-finite clause, next to the noun being modified, and that they then inflect like adjectives. If the non-finite clause is an infinitive clause, it postmodifies the noun in German, eg "die Fähigkeit, das Klavier zu spielen" 'the ability to piay the piano'. Since, unlike relative clauses, non-finite clauses generally lack a sentence element, viz the Subject, the relationship of the modified noun to the clause is not made explicit. In fact, the modified noun is the implied Subject of the non-finite clause, at least of participle clauses: in the case of a present participle clause, the Subject of an active clause ("the girl looking at the book"—"the girl is looking at the book"); and in the case of a past participle clause, the Subject of a passive clause ("the accident witnessed by the people at the bus stop"—"the accident was witnessed . . ."). The relationship of a noun to a modifying infinitive clause is more problematical. In "the person to see about your problem", the modified noun person would appear to be the implied Object of the infinitive clause, cf "(you) see the person about your problem". But in "the first person to discover oxygen", the Expanding the noun (2) 73 modified noun person is implied Subject in the infinitive clause ("the person discovered oxygen"). And in "the ability to play a musical instrument", the modified noun ability has no recognisable implied function in the infinitive clause: ability is derived from the adjective able, or rather from the verb be able, so that the noun phrase is perhaps relatable to the sentence "someone is able to play a musical instrument". Exercise 29 In the following data (SIL 1980: Gil) based on Agatu (Nigeria), describe the possessive phrase and the relative clause: 'The child came' 'The dog came to the compound yesterday' 'Father came yesterday' 'The child saw the dog yesterday' 'The father of the two children came yesterday' 'The child saw the father's dog' 'The child saw two dogs' 'Father saw the child's two dogs' 'The child saw the compound' The compound-head (father-of-compound) saw the child's pot* 'The dog saw the compound-head's child' 'The compound-head's two dogs came yesterday' 'Father saw the child who came yesterday' 'Father saw two dogs of the child who came yesterday' 'The child saw the two dogs that came yesterday' 'The child who came yesterday saw two pots' 'The child who saw the compound-head's two pots came yesterday' 1. oi wa child came 2. ewowa ole ene dog came compound yesterday 3. ada wa ene father came yesterday 4. oi ma ewoene child saw dog yesterday 5. adagoiepa waene two 6. oimaewogada 7. oimaewoepa 8. adamaewoepagoi 9. oimaols 10. adagolemaehigoi pot 11. ewomaoigadagole 12. ewo epa gada gale wa ene 13. adamaoini waene who 14. ada ma ewoepa got ni waene 15. oimaewoepaniwaene 16. oi ni waene ma ehiepa 17. si ni ma ehi epa gada gole wa ene 14. Expanding adjective and adverb Not all languages, as we noted in Chapter 12, have separate classes of adjective and adverb (cf the Apinaye data in Exercise 5). Or, as in German, there may be considerable overlap between identifiable classes of adjective and adverb. Nevertheless, it is often possible to identify different functions corresponding to those of the adjective and the adverb in English, ie modifier of nouns (cf Chapter 12) or Complement (cf Chapter 11) for adjectives, and Adjunct (cf Chapter 11) for adverbs. In this chapter we want to look at ways in which adjectives and adverbs may be expanded. The possibilities of expansion are severely limited by comparison with those for nouns. Intensifiers In many languages, adjectives and adverbs may be expanded by a set of words known as intensifiers, themselves often regarded as a subclass of adverbs. The representative intensifier in English is very, eg in "very beautiful", "very quickly". In fact, very only ever functions as an intensifier in English, as does the equivalent word in French, trěs. In German, however, sehr may function alone as an Adjunct as well as in combination with an adjective or adverb as intensifier, eg "es regnet sehr" 'it's raining a lot', "sehr schön" 'very beautiful(ly)'. Not all words functioning as intensifier in English are restricted to that function, cf absolutely as intensifier in "His suggestion is absolutely marvellous", and as Adjunct in "I repudiate his accusation absolutely". In the Grammar of Contemporary English (Longman 1972), Quirk et al identify three classes of intensifier in English, in recognition of the fact that some 'in-tesifiers' have the opposite function to that indicated by their label; ie there is in fact a scale of intensification. The three classes of intensifier are those of: era-phasisers, amplifiers, and downtoners. It should be pointed out that some intensifiers in English have functions other than those of expanding adjectives and adverbs. The following examples will serve to illustrate the classes of intensifier in the expansion of adjective/adverb in English: emphasiser—"definitely shut"; amplifier—"completely clean", "thoroughly badly"; downtoner—"rather flat", "moderately important", "hardly new", "almost illiterate". It will be noted that intensifiers in English are generally placed before the adjective or adverb they modify, though enough is an exception to this rule, cf "bad enough". Not all adjectives and adverbs in English may be expanded by means of an intensifier. In the case of adjectives, it is the so-called 'gradable' adjectives that may be readily expanded in this way, ie adjectives with a 'more-or-less' meaning rather 74 Expanding adjective and adverb 75 than an 'either/or' meaning. For example, bad is a gradable adjective, since there are degrees of badness: something may be "more bad" or "less bad", or indeed "very bad", "completely bad" or "fairly bad". On the other hand, wooden is not a gradable adjective, since something either is made of wood, or it is not; so that we cannot talk of *"a very wooden toy", for example. Sometimes, though, intensifies are used with non-gradable adjectives, when they have an 'emphasising' function, cf the "definitely shut" example above, or the much criticised expression "very unique". In the case of adverbs, it is the set of conjunctive adverbs, like however, therefore, moreover and besides, which may not be expanded by means of intensifiers. Also, the set of pro-form adverbs—here, there, now, then—do not normally occur with intensifiers, though there are ways in English of 'emphasising' these adverbs, eg "righthere", "just then". To illustrate the intensification of adjectives in another language, consider the following data (based on SIL 1980: D8) from Tataltepec (Mexico): 1. ngatsi late? bP 'That cloth is yellow' yellow cloth that 2. ngatl Ji kiňa?älate? 'Many cloths are fairly white' white fairly many 3. tljuľinatí bP 'That person is fairly large' big person 4. tljutsa: ska ^uru 'One donkey is very big' big very one donkey 5. tljutsa: 'It is very big' 6. ngati tsa: late?tlju re 'This big cloth is very white' this 7. ndPí tsa: ňatí 'There were lots of people' there-was lots Most of the sentences in this data are basea on a C-S pattern, without a Predicator. Adjectives like ngati 'white', tiju 'big' may be expanded by intensifiers such as/;' 'fairly' and tsa: 'very', which are placed after the adjectives they modify. Note that tsa: also has a function as a quantifier in the expansion of a noun (No 7), ie as a member of the same class as ska 'one' &x\dkiňa?ä 'many'. Comparative expressions Another way of expanding adjectives and adverbs is by means of comparative expressions. This kind of expansion is possible with the same sets of adjectives and adverbs that may be subject to intensification. Comparative expressions involve on the one hand either an appropriate (eg comparative) form of the adjective or a comparative adverb, and on the other some kind of comparative phrase or clause; eg "worse than we had imagined'', "the most beautiful jewel in the whole world". In English, 'more/less' comparison has either a noun phrase introduced by the preposition than or a clause introduced by than as a conjunction, eg "bigger than an 76 Discovering grammar elephant", "less frequent than the timetable states". Superlative expressions; ('most/least' comparison) have either a locative expression or a clause introduced^ by conjunction that, eg "the least expensive in the shop", "the most reliable that" money can buy". There is a similar expression in English with too and an infinitive clause, eg "too tired to walk another step"; and also a comparative expression of 'equality' with as either as preposition or as conjunction, eg "as foolish as the next man", "as foolish as I thought he would be". All the examples so far have been with adjectives. Similar expressions occur also with adverbs, eg "more gracefully than í have ever seen", "most patiently of all the prisoners", "too decisively to be ignored", "as quickly as he was able". And similar expressions can be found in other European languages at least; eg French "plus fort que son frěre" 'stronger than his brother', "plus vite que je ne le croyais" 'faster than I thought' (note the insertion of ne 'negative particle' and the pronoun le 'it' in the comparative clause); German "so kalt wie es im Januar war" 'as cold as it was in January', "zu müde, einen weiteren Schritt zu machen" 'too tired to take another step'. In Punjabi there are no comparative or superlative forms of the adjective as in English or German. Comparison relies on the use of the comparative particle equivalent to English than: tor) or nahotj. For example: "mwn,c|a kwri torj bmmae" 'The boy is taller than the girl' boy girl than tall is "omsythorjsyartee" 'He is cleverer than me' he me-than clever is With the superlative, the quantifying pronoun sab 'all' is used with the comparative particle; eg "é kar dassbtorj csngaksmrae" 'This is the best room in the house' this house of all than good room is Note that the phrase in this last example is a possessive phrase (kar do) rather than a locative phrase as in English, cf German "das grösste Zimmer des Hauses", lit. 'The biggest room of the house'. In the Hixkaryana language of Brazil (cf Desmond C Derbyshire, Hixkaryana, North Holland 1979), there are three ways of expressing the comparative relation. The first is by j uxtaposing a negative and a positive clause! eg "kawohra naha Waraka. kaw nahaKaywerye" tall-neg he-is tall he-is which could be translated either as 'Waraka is not as tall as Kaywerye', or as 'Kaywerye is taller than Waraka'. The second is by juxtaposing positive clauses containing 'degree' particles like nyhe 'more', rmahaxa 'very much': eg "ohxe naha meku. ohxe nyhe naha yayhi.ohxe rmahaxa naha honyko" good it-is monkey more tapir very-much peccary ie 'Monkey is good, tapir is better, and peccary is really good'. Expanding adjective and adverb 77 The third means is by the use within the clause of postpositional relators -oho 'bigger than, more important than' and -osnaka 'less than, smaller than': eg"kratxatxa yoho nahatukusu" grasshopper bigger-than it-is humming-bird 'The hummingbird is bigger than the grasshopper' "enahri yoho rmahaxa tinyahke natxow hami" eating-of-it greater-than very-much having-food they-are 'deductive' 'It is evident that they have much more food than they can eat'. Adjectives—phrasal/clausal expansion One further way of expanding adjectives—in English at least, though it is probably by no means universal—is by means of a phrase or a clause. The type of phrase concerned is a prepositional phrase (ie noun phrase introduced by a preposition), eg 'fond of cream cakes", "anxious about the future", "annoyed at the suggestion". It will be noted that particular adjectives are associated with specific prepositions; compare further "keen on ice hockey", "different from his brother", "disgusted with the exam result", "contrary to all advice", "eager for new experiences". With some adjectives the noun phrase which the preposition introduces may be replaced by an -ing-clause (clause introduced by a present participle form of the verb) or by a w/i-clause (introduced by a lwh' word, eg what, who, when), eg "guilty of robbing the bank", "uncertain about what they should do next", "angry about where the telephone had been put". The clausal expansion of adjectives in English is by means of a f/jaf-clause or an infinitive clause, eg "concerned that nothing would be done", "anxious to arrive at work on time", "eager to please everybody", "important that nothing should be forgotten". Adjectives that are expanded in this way, and indeed by means of a prepositional phrase, may only function predicatively in English, usually after the verb be. It is arguable that 'be + adjective' should thus be regarded as a verb, which is what the equivalent would be in some languages, cf be afraid—German sich fürchten. Exercise 30 Describe the expansion of adjectives in German from the following data. Note that German has a four-term case system: nominative (nom), accusative (ace), genitive (gen) and dative (dat): 1. Er war seinen Eltern gehorsam 'He was obedient to his he was his(dat) parents obedient parents 2. Wir wollen unserer Sorgen iedigsein 'We want to be free of our we want our(gen) troubles free to-be troubles' 3. Er ist zu dieser Aufgabe fähig 'He is capable of this task' he is to this(dat) task capable 4. Sie ist gegen dieses Medikament empfindlich 'She is allergic to this she is against this(acc) medicine allergic medicine' Discovering grammar 5. Dieses Haus ist nicht den Preiswert this(nom) house is not the(acc) price worth o. br ist es müde, dauernd zuarbeiten he is it tired continually to-work 7. Sie ist von seiner Unschuld überzeugt sheis of his(dat) innocence convinced 8. Sie ist auf die Leistung stolz sheis on the(acc) achievement proud 9- Er ist für den Schaden haftbar cClS for the(acc) damage responsible iU. Er ist damit einverstanden, dasswir sein Autoboxen he .s it-with agreeable that we his(acc)car borrow 'This house is not worth th« price' " 'He is tired of continually working' J I 'She is convinced of his innocence' 'She is proud of the achieve ment' 'He is responsible for the damage' 'He is aggreable to us borrow. inghiscar' 15. Locatives and temporals Two of the most common kinds of circumstantial information that we choose to specify for any event are 'where' the event took place and 'when' it took place; that js 'locative' information and 'temporal' information, respectively. Location and time are, however, not simple concepts. In the case of location, for example, one may specify the place 'ať which an event takes place, or the place 'towards' or ■from' which movement occurs, besides the position of things relative to each other (eg 'in front of, 'behind', 'beside', 'below', etc). Similarly in the case of time, one may specify the point in time 'ať which an event occurred, or the amount of time that an event took up (duration), or the relative distribution in time of events (eg 'before', 'after', 'while'). Another fairly common kind of circumstantial information that is specified is the Manner in which an event takes place or an action is carried out, answering the question "How did it happen?" or "How was it done?". Other kinds of circumstantial information include: purpose, cause, reason, result, condition, etc. Such information usually occupies the Adjunct slot in sentence structure (see Chapter 11). In this chapter we shall consider the various grammatical means for expressing circumstantial information, in particular locative and temporal in-[ formation. Adverb phrases Adverb phrases are relatively abbreviated forms of locative and temporal expression, consisting often of only a single adverb without intensification. Sometimes, especially in the case of locatives, they have a pro-form function, referring back in a text to a fuller expression of location or time by means of pre/postpositional phrases or adverbial clauses (see below). This is particularly true in English, for example, of the adverbs here, there, now and then. Other locative adverbs in English include the adverb particles like in, out, below, above, behind, as in sentences like "Push the knob in", "We'll clamber below", "Another bus is coming behind". Besides now and then, other point-of-time adverbs in English include today, yester-day (though in some contexts these might be considered to be nouns), soon, just, afterwards, before. There' are also time-frequency adverbs such as: often, sometimes, occasionally, frequently, usually, again, never. Most commonly expressed by adverb phrases, in English at least, is Manner: all the adverbs formed from adjectives by the addition of the derivational suffix -ly are manner adverbs, eg &acefully, quietly, happily, beautifully, slowly, candidly, as in "She danced gracefully'', "They moved quietly", "The children play very happily together". 79 80 Discovering grammar Consider now the adverb phrases in the following data (SIL 1980: E2), based on Northern Pueblo Totonac (Mexico): 1. wal he-spoke 2. tj"u?ncanawan thus he-will-speak 3. wai maqiuwa many-times 4. tJ"o?la watlakapal probably quickly 5. tju^ncawanqa'wa^tjulaqalinawan the boy tomorrow 6. nawan maqiuwa laqali tuJuma'n 'He will speak many times in the future' day-after-tomorrow 7. pedrowailaqa:tfu Peter everywhere 8. qo:tan pedroki:wai wampala yesterday went-to-speak again 9. a:kalista:n nawan pedrolakapai after-that 10. nawan laqa:tj"uka:cisa:t early 11. wal a'ca maqiuwa here The locative adverbs in this data are: laqa.tfu 'everywhere', a?ca 'here'. The: temporal adverbs are, firstly time-when: laqali 'tomorrow', tu/uma^ 'day-after-tomorrow', qo.tan 'yesterday', a:kalista:n 'after-that', ka:cisa:t 'early'; secondly time-frequency: maqiuwa 'many times', wampala 'again'. The manner adverbs are: tfu^nca 'thus', lakapat 'quickly', tfo^la must also be considered as an adverb, perhaps belonging to a class of 'modal' adverbs (cf Chapter 17). Prepositional/ postpositional phrases We have encountered pre/postpositional phrases already, eg in the expansion of adjectives (Chapter 14), and in the expansion of nouns (Chapter 13). Here wft consider them functioning as Adjunct with locative, temporal and other circum-: stantial meanings. In 'prepositional' phrases the particle precedes the associated, noun (phrase), eg English "in the garden"; whereas in 'postpositional' phrases the| particle follows the associated noun (phrase), eg Punjabi "bazar yc" (lit. 'bazaaxj in') 'in the bazaar'. Not all languages have pre/postpositional phrases (eg Turkish)| but rely instead on a system of cases in the noun, so that nouns or noun phrases iß. the appropriate case signal locative/temporal information. Other languages (eg Latin, German) have a system of prepositions governing the case of the associated noun; eg German "in die Stadt" (in + accusative) 'to the town', "in der Stadt" (ôi + dative) 'in the town'. Locatives and temporals 81 pre/postpositional phrases as Adjunct express primarily locative and temporal information. Examples of locative prepositional phrases in English are: "The incident occurred outside the swimming baths", "The children are playing in the garden", "Grandpa is asleep on the sofa". And examples of directional phrases: "We're going into the town centre", "They raced across the park", "We've just come from the cricket match". Here are now some examples of temporal prepositional phrases in English, firstly time-when phrases: "The incident occurred in the morning", "We all went to sleep after our lunch"', "The manager can see you at any time during the day". Secondly, time-frequency/duration phrases: "They made the noise throughout the afternoon", "The voyage lasted for six days". Time-duration is also expressed by noun phrases, without preposition, in English; eg "The performance lasted four hours", "They were at sea ten days". Time-when is also expressed in this way with days of the week and the like, eg "The accident happened last Friday", "We shall go to Greece next Spring". Prepositional phrases perform a similar function of expressing locative/temporal information when they occur in the expansion of nouns, eg "the house on the corner", "the discussion after lunch", "my thoughts during the lecture". But prepositional phrases as modifiers of nouns do not always perform this function; sometimes a noun requires a particular preposition + noun (phrase) as a kind of complement, eg "his request for asylum", "her annoyance at the rumours", where for and at are not replaceable by any other preposition but are determined by the nouns request and annoyance, respectively. Exercise 31 Describe the locative and temporal expressions in the following data (SIL 1980: E4) from Bekwarra (Nigeria): 1. a kä ye ká he then went there 2. a man ye ká again 3. a be kin came here 4. a kä man be 5. abeufufo yesterday 6. ayekáufufo 1- abeyě ká from 8- a kä ye yě kin I 9. ayek'utyenufufo h to farm 10- a ci kin yě ufufo i stayed *!• aneum ká ť saw someone P2- a kä manne he ká him 'Then he went there' 'He went there again' 'He came here' 'Then he came again' 'He came yesterday' 'He went there yesterday' 'He came from there' 'Then he went from here' 'He went to the farm yesterday' 'He has stayed here since yesterday' 'Hesaw someone there' 'Then he saw him there again' 82 Discovering grammar Temporals and the verb phrase Clearly, temporal expressions, particularly referring to time-when, frequently ha\S the function of amplifying or making more specific the tense marked in the vem phrase. So, in the sentence "They arrived yesterday", the -(e)d of arrived indicate* the point in time at which the event occurred as being past, and the temporal adverb yesterday makes that point in time specific. In Urdu, the word for 'yesterü day' is the same as the word for 'tomorrow'; the distinction is provided by the tense of the verb. Compare: "kal kya dyntha?" 'What day was it yesterday?7 yesterday which day it-was "kal kyadynhoga?" 'What day will it be tomorrow?' tomorrow it-will-be Likewise, the same word is used for 'the day before yesterday' and for 'the day after tomorrow', V\zpsrsox]. The reverse situation obtains in English with some expressions of future time^ where the burden of future reference rests solely with the temporal expression; el in "They are visiting the museum tomorrowf on Friday j next week", where the tense is 'present progressive1, and the absence of the temporal expression wouff remove reference to future time. A similar dependence on the temporal expression occurs with the simple present tense, eg "They visit the museum tomorrows though here the removal of the temporal expressions creates a gap that is appropriately filled by a time-frequency expression, eg "They visit the museum every. week/on Wednesdays". Time-frequency meanings in English may be expressed within the verb phrase withj out the presence of a temporal expression. Compare: "He always bangs the door when he comes in", with the temporal adverb always, and "He will bang the dooj when he comes in", where a similar meaning is expressed by the modal auxiliary verb will. A similar expression is the use of keep on in sentences like "He kept on banging the door", in which keep on could be regarded as some kind of auxiliary verb. Consider also the meanings of frequency and repetition expressed in different verb forms in the following data (from SIL 1980: A23) from Tlingit (Alaska): 1. nagúttj 'he goes frequently' 2. a xsattntj 'he sees it often' 3. askóowtj" 'he always knows' 4. yoo sya.átk 'they go back and forth' 5. a daa yoo toowatánk 'he's thinking something over' 6. at únt 'he's shooting (repeatedly)' 7. a klaxutt 'he's splitting (firewood)' 8. datátjt 'he's clapping (in time to music)' Adverbial clauses There is a further, more expansive means of expressing locative, temporal an3 indeed all kinds of other circumstantial information: adverbial clauses. Like adverb I Locatives and temporals 83 phrases and pre/postpositional phrases they can be regarded as having the function 0f Adjunct in sentence, structure, though in this case it is a matter of 'embedding' one clause in another clause (see Chapter 19). Adverbial clauses are frequently introduced by conjunctions, which specify the circumstantial meaning of the clause, eg in English where for locative adverbial clauses ("The accident happened where (he two roads meet"), when for temporal point-of-time adverbial clauses {"When you come home I shall have the meal ready"). Temporal adverbial clauses may be particularly important for indicating the time relationship between two events, eg by means of adverbial clauses introduced by the conjunctions before, after, while {"Before they went to bed they washed and brushed their teeth", "They went to bed after they had washed and brushed their teeth", "While they were brushing their teeth the telephone rang"). This is not the only clausal means of indicating the temporal relationship between events, however, neither in English nor in other languages: a participle clause, without a conjunction, may serve a similar purpose, eg "Having washed and brushed their teeth, they went to bed", "Brushing their teeth, they heard the telephone ring". There must, however, be an identity of reference between the implied Subject of the participle clause and the actual Subject of the main clause (cf "Brushing their teeth, the telephone rang"). Consider now the following data (from SIL 1980: F13) from Xavante (Brazil), where the temporal relator is not a conjunction, but a suffix in the verb of the adverbial clause: 1. tetemadaa-wamhä te mo 'When he had watched it, he he watched-it-when he went went' 2. teteupra-re wa mada T watched while he bought it' he bought-it-while I watched 3. te mahari-mono-re teteteupra 'Whenever I made it, he bought I made-it-habitually-while he bought-it- it' habitually 4. teteupsö-mono-wamhä tete madaa 'Whenever he washed it, I he washed-it-habitually-whenl watched watched' Note that the adverbial clause appears always to take up initial position in the sentence. Adverbial clauses may serve to express several further kinds of circumstantial information, besides locative and temporal. For example: cause or reason, introduced by because or since in English; condition, introduced by if oz negatively by unless; concession, introduced by although; result, introduced by so that; purpose, introduced by in order that or simply an infinitive clause ("Jack and Jill went up the hill to fetch a pail of water"). Eg "He arrived late, because his car broke down", "If you agree to my suggestion, I will help you", "The museum is shut, so that we cannot visit it today". Exercise 32 Describe the adverbial clauses in the following Finnish data (SIL 1980: Fl): 1- Ostin auton 'I bought a car' I-boughtcar 84 Discovering grammar 2. rikastuin 3. elin rikastunut 4. rikastuttuani ostin auton 5. kun elin rikastunut ostin auton when I-had 6. rikastun 7. jos rikastun ostan auton if I-buy 8. menin kaupunkiin I-went to-town 9. menin kaupunkiin ostamaan auton to-buy 'I became rich' 'I had become rich' 'Having become rich, I bought a car' 'When I became rich, I bought a car' 'I am becoming/will become rich' 'If I become rich, I will buy a car' 'I went to town' 'I went to town to buy a car' 16. Agreement Items in a linguistic unit that are in some kind of syntactic relationship, eg Subject and Predicator in a sentence, adjective and noun in a noun phrase, may mark that relationship in some way. We noted, too, in previous chapters how, in German, prepositions are followed by nouns in a specific case, eg von 'of, from' by the dative, für 'for' by the accusative (see further below). Items that are in a syntactic relationship and have that relationship marked are said to be 'in agreement'. In this chapter we shall consider the kinds of syntactic relationship that are marked and the means employed for marking them, ie the kinds of 'agreement' that occur in languages. Noun phrase Items in the noun phrase—determiners, adjectives, nouns—are frequently found to be 'in agreement'. Consider the English noun phrase "these girls": by comparison with the singular form "this girl", it can be seen that both demonstrative (this— these) and noun {girl—girls) are marked for 'plural' number. There is agreement between demonstrative and noun in respect of number, marked (in the spoken form) in the demonstrative by the lengthening of the vowel (i-»i) and voicing of the final sibilant (s-»z), and marked in the noun by the addition of the suffix /z/. This is probably the only kind of agreement that can be found in English noun phrases. Consider now the following examples of French noun phrases: 1. lebongarcon 'the good boy' 2. lesbonsgargons 'the good boys' 3. la bonne fille 'the good girl' 4. les bonnes filles 'the good girls' Taking the written forms for our description—the description of the spoken French forms would be much more complex—it is clear that there is agreement between article, adjective and noun in respect of 'plural' number (Nos 2 and 4), marked by the -s suffix on each item (additionally there is a change of vowel in the article in No 4, as compared with No 3, from a toe). In addition, there is agreement between the three items in respect of gender (gargon is a 'masculine' noun, fille is a 'feminine' noun), marked in the article by the alternation between le and la, marked in the adjective by the addition of -ne for feminine gender, but not marked in the noun, which may be regarded as having inherent gender. In Nos 2 and 4, then, we may say that there is agreement in respect of both gender and number between article, adjective and noun, marked in the ways indicated. When we turn to German we find agreement in the noun phrase, not only in respect °f number and gender, but additionally in respect of case. The German noun 85 86 Discovering grammar (phrase) has a three-term gender system ('masculine', 'feminine' and 'neuter') agl a four-term case system ('nominative', 'accusative', 'genitive', 'dative'), Consili the following paradigms for the noun phrases "der gute Mann" 'the good manff "die gute Frau" 'the good woman', and "das gute Boot" 'the good boat': Masculine Feminine Neuter sing • pi < nom acc gen dat nom acc gen dat der gute Mann den guten Mann des guten Mannes dem guten Mann die gute Männer die guten Männer der guten Männer den guten Männern die gute Frau die gute Frau der guten Frau der guten Frau die guten Frauen die guten Frauen der guten Frauen den guten Frauen das gute Boot das gute Boot des guten Bootes dem guten Boot die guten Boote die guten Boote der guten Boote den guten Booten It will be noted that there is by no means maximal differentiation of forms; there i&J in fact considerable overlap, eg the -en form of the adjective (guten) appears! throughout the plural paradigms and in a good number of cases in the singularj ones. Similarly the article form die occurs in nominative and accusative in all threej genders in the plural, in addition to the feminine singular in those two cases. Hj should, however, be pointed out that the form of the adjective depends on the kindj of determiner present, and whether a determiner is present or not, cf "ein guter] Mann" (nominative singular), "guter Männer" 'of good men' (genitive plural). Aj principle of economy appears to be at work in these paradigms, requiring the clear, marking of gender, number and case once only in each noun phrase. J Consider now the following data (SIL 1980: C2) from Basari (Ghana), set as Exer^ eise 7in Chapter 3: 1. uni umbini 'this person' biniib bimbini 'these persons' 2. uboti umbini 'this chief bibotiib bimbini 'these chiefs' 3. diyin dimbini 'this name' ayin rjimbini 'these names' 4. dibil dimbini 'this seed' abil rjimbini 'these seeds' 5, kusaau kumbini 'this farm' tisaati timbini 'these farms' 6. kukabuu kumbini 'this basket' tikabuti timbini 'these baskets' As we saw in Exercise 7, the noun and the demonstrative are marked by means of prefixes and suffixes according to gender (1, 2 and 3) and number (singular and plural) in the following way: Noun Demonstrative Gender I -bsti 'chief -ni 'person' sing -. Pi u-bi- -ib Gender 2 -yin 'name' -bil 'seed' sing P' di-a- Gender 3 -íťjíi-'farm' -kabu- 'basket' sing Pi ku--u ti--ti u- bi- di-rji- ku-ti- Agreement 87 That is to say: there is agreement between noun and demonstrative in respect of gender and number, marked by affixes as shown in the table above. exercise 33 Describe the agreement in the Portuguese noun phrase from the following data (SIL 1980: C3). The order of elements is: demonstrative—noun—adjective, 1. esse menino gordo 'that fat boy' 2. essamenina bonita 'that pretty girl' 3. esses meninosgordos 'those fat boys' 4. essasmeninasbonitas 'those pretty girls' 5. esse vestidobonito 'that pretty dress' 6. essagalinhagorda 'that fat hen' 7. esses vestidos bonitos 'those pretty dresses' 8. essasgalinhasgordas 'those fat hens' Sentence We turn now to agreement within the sentence. Here the agreement usually involves the marking of a syntactic relationship between the Predicator and some other sentence element, often the Subject. Compare the following pair of sentences in English: "My son plays chess"—"My sons play chess". The difference between singular and plural is marked in both the Subject and the Predicator, in the singular by the absence of -s in the Subject noun with corresponding presence of -s in the verb, and in the plural by the presence of -s in the Subject noun with corresponding absence of -s in the verb. Thus, we can say that there is agreement between Subject and Predicator in English with respect to number, marked by the presence/absence of an -s suffix as outlined above. However, this agreement applies only to 3rd person Subjects and only in the presence tense, except for the verb be (cf "I am/you are/he is", "I/he was/you were"). In written French, on the other hand, as we saw in Chapter 4, there is consistent agreement in respect of person and number in all tenses, marked by suffixes in the verb, with little duplication of forms. A similar extensive agreement obtains between Subject and Predicator in German; cf the following present tense paradigm for gehen 'go': Singular Plural 1st person ich gehe wir gehen 2nd person du gehst Ihr geht 3rd person er/sie/es geht Sie/sie gehen In Gujerati transitive sentences the Predicator agrees with the Object, in some instances in respect of gender, number and case, in others in respect of gender and number only. For example (from W S Tisdall, Gujerati Grammar, New York: Ungar 1961): 1. "E baie vagaclamaň dahad^a kahacjja", 'That lady passed her days in the wilderness'. Here kaha<\ya 'passed' is in the 'absolute' case, 'plural' number, 'masculine' gender, to agree with daha<[a 'days'. Ď0-D 88 Discovering grammar 2. 3. "Tene e rajae pakaciyo", 'The king seized him'. Here pakaciyo 'seized' is mas| culine singular, to agree with tene 'him', but it is in the 'absolute' case while tene if in the 'oblique' case with -ne. "Ten^e Ranjne ranmadaň nasadj muki", 'He removed the queen into a forest*. Here the verb muki is in the feminine singular, agreeing with the Object Ratf 'queen', though Ratii is in the 'oblique' case with -ne. Consider now the following data (SIL 1980: Africa/Mozambique): C9) from Tsonga (South 1. mufana watlarjga 2. mufana lonkulu watirha 3. mufana wareka 4. mufana watsutjuma 5. mufana wadja 6. mufana wadja Jihari lejikulu 7. ßafana ßatlarjga 8. ßafana laßakulu ßatirha 9. ßafana ßaleka 10. ßafana ßatsut Juma 11. ßafana ßadja 12. Jihari Jatsutjuma 13. Jihari le fikulu Jabaleka 14. sihari satsutjuma 15. sihari lesikulu sabaleka 'The boy plays' 'The big boy works' 'The boy laughs' 'The boy runs' 'The boy eats' 'The boys eats the big animal' 'The boys play' 'The big boys work' 'The boys laugh' 'The boys run' 'The boys eat' 'The animal runs' 'The big animal runs away' 'The animals run' 'The big animals run away' In this data there is agreement both within the noun phrase (Nos 2, 6, 8, 13, 15) and within the sentence. Agreement in the sentence is between Subject and Predi-cator with respect to gender and number, marked by prefixes in the Subject noun and in the verb as follows: Subject Predicator Gender 1 •fana 'boy' Gender 2 -hari 'animal' sing PL sing pl mu-ßa- Ji-si- wa-ßa j> sa- The agreement in the noun phrase is between noun and adjective, with respect to number and gender, marked by prefixes as follows: Noun Adjective Gender I Gender! sing Pl sing pl mu-ßa- Ji- si- lon-laßa- leji-lesi- (No2) (No 8) (Nos 6, 13) (No 15) Agreement 89 Concord and government Sometimes in the discussion of agreement a distinction is made between two kinds of agreement: 'concord' and 'government'. The distinction refers to the way in which agreement is marked. In the case of concord, all the items in agreement are marked in some way. So, in the Tsonga data just considered, both Subject and Predicator are marked by prefixes in the sentence agreement, and in the noun phrase agreement both noun and adjective are marked by prefixes. So, these are both instances of concord. Government, then, refers to the type of agreement where not all items are marked: one item, the unmarked one, is said to 'govern' the form of the other items in the syntactic unit. In the French singular noun phrase, for example (cf "la bonne fille", 'the good girl'), the noun is not itself marked for gender (though it belongs to the 'feminine' gender), but it governs the form (feminine) of the article la and the adjective bonne. Things are complicated in the plural noun phrase, however, because 'plural' number is marked in all the items in the unit (cf "les bonnes filles"). Perhaps a too rigid insistence on the distinction between concord and government will in some cases be descriptively unhelpful. How, for example, could one fit the agreement in English "The boy plays"/"The boys play" into this distinction? Pre/postpositions As we have seen before (cf Chapter 15) prepositions or postpositions are often associated with nouns in a particular case. This could be considered a form of government: a pre/postposition governs the case of an associated noun. Sometimes pre/postpositions in a language all govern the same case; eg in Punjabi, postpositions are preceded by nouns in the 'oblique' case, or in English if a preposition governs a personal pronoun it is in the 'object' case (cf "to me", "for him", "from us"). In other languages (eg German) the situation is more complex. Some prepositions in.German always govern the same case, eg von ('from', 'of) + dative, für ('for') + accusative, wegen ('because of) + genitive. The majority of prepositions are of this kind. A few prepositions, however, may govern more than one case, usually with a difference of meaning; eg in ('to', 'towards', 'into') + accusative, in ('in', 'inside') + dative; au/('onto') + accusative, auf'('on', 'on top of) + dative. Exercise 34 Describe the agreement in the following data (SIL 1980: CIO) from Mbembe (Nigeria): 1. oni ope „ 'The elephant died' elephant died 2. ojinorjope The man died' man 3. eten eze ojinorj The animal saw the man' animal saw 4. ajinorj maze eten ndoma The men saw this animal' men saw this 90 Discovering grammar 5. ojinonrjwomaoze nten nta:n this saw animals three 6. ajinorj mboma ata:n maze nten njoma nta: these three these 7. ajinorj ata:n mape died 8. nten njoma nta:nipe died 9. oňi rjwoma ope za there 10. aňi mboma ata:n mape za elephants 11. ojinorj rjwoma oze ani mboma za 'This man saw three animals' 'These three men saw these three animals' 'Three men died' 'These three animals died' 'This elephant died there' 'These three elephants died there' 'This man saw these elephants there' 17. Mood and voice In this chapter we consider two grammatical categories that are commonly associated with the verb phrase, but which also often have implications for sentence ordering and structure. The category of 'mood' is frequently understood to involve two separate sets of distinctions, on the one hand the contrast between declarative/affirmative, interrogative, negative, imperative and perhaps subjunctive, and on the other hand the expression of notions such as necessity, probability and certainty, and of such notions as obligation, ability and volition. The category of 'voice' refers primarily to the distinction between 'active' and 'passive', though some languages distinguish also a 'middle' voice. Interrogative mood In discussing the interrogative and other such moods, the assumption is usually made that the declarative or affirmative mood is somehow basic, and interrogative, negative, etc, sentences are derived from declarative ones. Mood is sometimes referred to as 'sentence modification'. Essentially, moods concern the role which a speaker takes on in relation to a proposition, ie as questioner, requestor, commander, denier, etc. Languages vary in the ways in which moods are realised grammatically, whether morphologically or syntactically. It is useful to distinguish two kinds of interrogative sentence: 'polar' questions and 'information' questions. Polar questions ask merely about the polarity (negative/positive) of a proposition, and so they expect the answer "Yes" or "No", eg in English "Have you let the cat out?" Information questions, on the other hand, ask for some specific information and include an information-seeking word (interrogative pronoun, adverb or adjective) which identifies the specific information, eg in English "Who is your brother talking to?", "Where did you put my scissors?" Here who is an interrogative pronoun that expects a noun (phrase) for an answer, and where is an interrogative adverb that expects an Adjunct of some kind for an answer. In English, interrogative mood is realized largely by syntactic rearrangement. This is certainly true in the case of polar questions, which are formed from declarative sentences by the inversion of Subject and first auxiliary verb in the verb phrase, eg "You have let the cat out" -* "Have you let the cat out?" If there is no auxiliary verb in the verb phrase of the declarative sentence, then the dummy auxiliary do is used in the question, eg "You gave Susan the book" -* "Did you give Susan the book?" In information questions in English the information-seeking word is placed first in the sentence, followed by the first auxiliary verb, then the Subject, the remainder of the verb phrase and the rest of the sentence, eg "My brother has been talking to the headmaster" -*• "Who has my brother been talking to?" Here the 91 94 Discovering grammar by means of a declarative sentence, eg "I'd rather like the newspaper". For th" reason it is probably important to distinguish mood (as a grammatical category from sentence function (as a semantic/pragmatic category), since they are n t necessarily in a one-to-one relationship. In the subjunctive mood a speaker casts doubt on the validity or veracity or realitv of the proposition he is making. In English, the subjunctive survives only in a few fossilised phrases like "If I were you . . ."; its function has been taken over by the modal verbs should and would. In French, the subjunctive verb form is found in subordinate clauses after certain verbs of cognition, eg croire 'believe', vouhir 'want/'wish': "Je crois qu'elle soít morte" ('I believe her to be deaď) "Je v«,* qu il vienne ( I want him to come'). The subjunctive is also associated with other moods or roles that a speaker may adopt, eg wishing, cf English "Long live the Queen", "God bless you", where the verb form lacks the usual -s suffix of the 3rd person singular present. With the subjunctive and its expression of doubt and uncertainty we have come close to the other set of distinctions under the heading of 'mood', sometimes also called 'modality', to distinguish it from the set we have so far considered. Exercise 35 Describe the signalling of interrogative mood in the following data (SIL 1980" F9i from Yakurr (Nigeria): " " ' " 1. odarn owuokohkebla 'Your husband has gone to the farm' husband your go farm 2. odamowuokobkebla-o 'Has yourhusband aone tothefarm?' 3. umana-wahmoyahu kah 'My friend is there'" friend-my is there 4. umana-wahmoyahu kah-o 'Is my friend there *?' na yakehwyatietehn , They are going to kill an animal' future they-go kill animal 6. nayakehwyatietehn-o 'Are theygoingto killan anima 7. wo! liwehlitahwa 'Heiswelľ body his is-strong 8. wolliwehlitahwa-o "Is he well?' Mood—modality We turn now to the second set of meanings associated with the term 'mood', but also called 'modality'. Again we are concerned with a role, or stance or attitude that a speaker takes up in relation to the proposition that he is making. In this case, however, the speaker intervenes with his assessment of the possibility, probability or certainty of the proposition. One way in which English expresses these kinds of meanings is by means of the so-called 'modal' auxiliary verbs: can. may, must, will, shall. For example, "The letter may come tomorrow", "It could come tomorrow", "It will come tomorrow", "It must come tomorrow" express varying degrees of certainty/uncertainty or possibility concerning the proposition. But the modal auxiliary verbs are not the only means of expressing these modal meanings in English: there is, for example, a set of modal adverbs (or modal particles) that may be added Mood and voice 95 to a sentence to modify the proposition in respect of its certainty or possibility: maybe, perhaps, for sure, certainly, possibly, probably, etc; eg "Perhaps the letter will come tomorrow", "The letter will come tomorrow for sure". Some of these modal adverbs also have adjective and noun counterparts, eg "It is possible that the letter will come tomorrow"—"There is a possibility that the letter will come tomorrow", "It is certain that the letter will come tomorrow". Thus English realises the meanings of modality in a number of ways. French, on the other hand, does not have the same range of modal auxiliary verbs as English, but more often uses adverbs and adjectives to express modality; eg "II vient demain peut-itre", 'He may come tomorrow'/'Perhaps he'll come tomorrow'. Consider now the following data (SIL 1980: E13) from Tlingit (Alaska), presented in a modified form as Exercise 25 in Chapter 11: 1. xóots shakdé saxwaa.áx dzeeyáak 'Maybe I heard a brown bear brown-bear voice-I-heard earlier-on earlier on' 2. Juneau-dé kwshé kukgwaatěen 'Perhaps he's going to Juneau he-will-take-a-trip on a trip' 3. kúnaxkusi.áať xáa 'It's sure cold' really it-is-cold 4. tlax a yáanax ee wdixwétl kwshé 'Maybe you're too tired' it face-beyond you is-tired 5. ee xôonee kwshé 'Maybe it's your friend' your friend 6. aatlěin dáanaa xáa doo jěewoo 'He sure has lots of money' lots money his hand-loc. The underlined items represent modal particles (adverbs or perhaps modal auxiliary verbs); they appear to be positioned before action verbs (Nos 1 and 2) but after state verbs (Nos 3 and 4), but they can also occur in verbless clauses (Nos 5 and 6). Modal verbs in English have another set of meanings (permission, ability, obligation) which are arguably not a matter of mood or modality, unless one views these meanings as extensions of the primary possibility/certainty meanings of the modais. The distinction has been made between the speaker or discourse-oriented uses of the modais (ie speaker's assessment of possibility or certainty) and the subject-oriented uses (ie permission, obligation, etc) as in: "Can you ride a bicycle?", "You may leave now", "We must report to the police tomorrow". Passive voice When we turn to 'voice' wé are considering not so much the speaker's assessment of or attitude to what he is saying, but rather the speaker's choice of the way in which he presents, within the sentence, the elements of his proposition. Voice, then, is concerned not so much with the speaker's relationship to his proposition or his interlocutor, but to the text of which the proposition forms a part, whether that text is a monologue or a dialogue. On the other hand, the term 'passive' has a further 96 Discovering grammar connotation, namely that the Subject of the sentence is being acted upon rath 1 than being the actor. ' e.r. In a passive sentence the Object of the active sentence becomes the Subject a f the Subject of the active sentence takes a subordinate position in an 'agenť-phra while the verb is put into the passive form (in English, be followed by past nľ' ticiple). For example, the English active sentence "A bus hit the car" transform* into the passive sentence "The car was hit by a bus", where the car is Object in th active sentence but Subject in the passive, and a bus is Subject in the active sen* tence but in the 'agent'-phrase "by a bus" in the passive. In English it is possible to" leave the Agent unexpressed, ie to omit the agent-phrase, eg "The drivers of th two vehicles were killed". The passive transform is thus only possible for active sentences that contain an Object, ie are transitive. But this is, in fact, not the whole story of the passive in English: see Exercise 36 below. The passive is particularly important in a language like English, where the order of sentence elements is relatively fixed, since the syntactic function (Subject, Object etc) is largely denned by position. In a language with a well developed case system' that distinguishes the syntactic function of sentence elements, a passive structure is not so important from a textual point of view. In German, for example, which does have a passive construction, the rearrangement of sentence elements can be achieved without recourse to the passive, eg "Mein Freund besuchte meinen Vater" ('My friend visited my father')—"Meinen Vater besuchte mein Freund" (lit. 'My father (accusative) visited my friend (nominative)*), ie 'My father was visited by my friend'. German differs from English also in that an 'impersonal' passive is possible that is not relatable to an active sentence with an Object, eg "Es wurde getanzt" (liť 'it was danced1), ie 'There was dancing'. In Punjabi the passive is used considerably less frequently than in English and normally in an agentless construction, eg "ethe paňjabi boli jandie e" (lit. 'here Punjabi spoken Passive is'), ie 'Punjabi is spoken here'. In tenses formed with the past participle (past, perfect, past perfect, future perfect) transitive verbs have a special construction (in the active) that is superficially equivalent to the English passive construction, as follows: the Subject is put in the oblique case and is followed by the postposition m 'by', the participle agrees in gender and number with the Object; eg "mwnde ne ponjabi sykkhi e" (Hi. 'boy by Punjabi learnt is'), ie The boy has learnt Punjabi*. Middle voice Some languages, eg Ancient Greek, have for some verbs in some tenses separate forms for active, passive and middle voice. If active voice implies as Agent doing an action, and passive voice implies a Patient undergoing an action, middle voice implies a person/thing acting of its own accord or in its own interest. The distinctions may be illustrated by the following English sentences: 1. Harry opened the door —active 2. The door was opened by Harry —passive 3. The door opened —middle. Mood and voice 97 In English' middle voice (if No 3 can be so termed) is not marked by a distinct form 0f the verb: the Patient/Undergoer of the action occurs as Subject (as in the pastve example—No 2), but the verb is in the 'active' form (cf No 1). In other languages (eg German, French) the verb in No 3 would be in a reflexive form, eg German "Die Tür öffnete sich" (lit. 'the door opened itself). This kind of reflexive has, however, to be distinguished from reflexives that refer to an action that an Agent does to itself, eg "Sie wäscht sich" ('She's washing herself), which must be regarded as active in voice. Exercise 36 Describe the passive voice in English as illustrated by the following sentences: [. They were arrested by the border police. 2. Everyone who reaches 100 is sent a telegram by the queen. 3. A telegram is sent by the queen to everyone who reaches 100. 4. The trial will be held next month. 5. It is understood that a statement will be made this evening. 5. This bed was slept in by Queen Victoria. 7. The garden has been walked over. J. Hundreds of people are killed every year in road accidents.