Bruce L. Liles University of Missouri, St. Louis AN INTRODUCTORY TRANSFORMATIONAL GRAMMAR Prentice-Hall, Inc., Englewood Cliffs, N.J.For Robert W. Ackerman @ 1971 by Prentice-Hall, Ine. Englewood Cliff, New Jersey. All vights rexeved. No part of tit book mney ibe reproduced in any form or by any marans without peauision in weting from the publisher, 13.502204.0 (©) 12 5u2286-% (P) Library of Congress Catalog Gare Nunnner: 70122388 ‘Finted in the United States of America, Current printing (last number): 19 18 17PREFACE ‘Ten years ago a person cowld read most of the important works in linguistics in a few months, and the scholar could afford to buy most of them for his library. Since that time research in all phases of linguistics has progressed so rapidly that scholars who have been keeping up with develop- ‘ments in their discipline are finding it diffioult ro read all the relevant works published each year. ‘The person wanting to hegin a study of Finguistics is overwhelmed by the amount of material available, much of it far too technical for him to understand. The present book is directed toward this person, and itis limited in scope to English transformational grammar. The student of transformational grammar needs both theory and applications. This book attempts to fuse the two. Also, it shows the reader that there are still many aspects of the English language that arc poorly understood and that are being investigated. ‘The exercises at the end of the chapters allow the reader to work with the material that has been and to use the techniques he has learned. T would like to express my gratitude to the people who helped me in writing this book: to David H. Grady, who suggested that I write it and who worked with me in the initial stages of production; to Frances Cassel and James Foster, who read dhe manuscript and offered useful suggestions; to Wanda Giles, who proofread the final copy. It would be impossible to list all the books and articles in journals that have affected my thinking; their influence is found on almost every page of this book. sssedCONTENTS PART ONE PHRASE STRUCTURE chapter | grammars of english ‘Traditional Grammar Structural Grammar ‘Transformational Grammar chapter 2 the structure of the sentence chapter 3 the auxiliary chapter 4 lexical features Divisions of a Transformational Grammar ‘Transitive and Inteansitive Restrictions Subject-Verb Restrictions Determiner and Noun Resttictions ‘The Lexicon PART TWO TRANSFORMATIONS 1 chapter 5 the negative transformation chapter 6 question transformations chapter 7 transformational processes Adverbial Movement Indirect Objects Imperatives chapter 8 the passive transformationvii contents PART THREE TRANSFORMATIONS 11 chapter 9 principles of transformational grammar chapter 10 compounding, deletion, and pro forms chapter II relative constructions chapter 12 sentences as noun phrases PART FOUR PHONOLOGY chapter 13 transcription chapter 14 phonological features | chapter 15. phonological features Il chapter 16 syntax and stress chapter 17 stress and pitch chapter 18 phonological rules Bibliography Indexpart one PHRASE STRUCTUREchapter one GRAMMARS OF ENGLISH TRADITIONAL GRAMMAR Until the sixteenth century, Latin was the primary language of scholarship in England and the rest of Western Europe. ‘The only grammars studied in English school, therefore, were Latin grammars, which were designed to sive Englishmen the skills needed to read, write, and sometimes converse in this Lingua franca of Western Europe. During the sixteenth and sevens teenth centuries, grammars of English began to appear. Since these carly grammars were studied exclusively as an aid to learning Latin, then Figorously followed Latin models, although the structure of Euglish ie que different from that of Latin For an cxample of differences between these languages, we can look at hhow they indicate such relationships as performer and receiver of action In Latin these relationships are expressed primarily by the forms of the words. In the sentence Puer virum videt, “The boy aces the man.” we know that puer, “boy,” is the one perforining the act and virum, “maa,” ss the one receiving it because there is no ending on puer, but -um is added te wir to give virum, The sentence would mean the same thing if the word order were reversed: Virum puer videt. To alter the meaning, one must change the forms of the words, not their orders hence, Vir pucrum vider means “The man sees the boy,” since vir has no ending and pucrum ends in-um, In contrast, by the sixteenth centuey English was relying exclusively on word order to indicate these relationships. ‘The sentence The boy sees the man shows that the boy is the performer because the word boy precedes the verbs the man isthe receiver, since man follows the verb. A change ie word order produces a change in meaning (The man sees the boy) or a sneaningless sentence (Sees the man the boy). Logically, grammar of Latin should discuss the forms of words, whereas one of English sould give ‘more emphasis to the order in which they are arranged. The purpose of the 34 chapter one carly grammars of English, however,.swas. uot, 10 provide an accurate de~ scription of the language, but rather to serve as a basis for the sty of Latin _ grammar, English word order was largely ignored, and the meaning of such 3s as actor and receiver was emphasized as a preface to Latin During the Renaissance, Latin was replaced as the language of scholarship by English and the other Western European languages. Because of this new role for English, some Englishmen by the late seventeenth and carly eighteenth centuries were greatly concerned with refining their language, They felt that Enelish had somehow become “corrupt” and that it was in need ot purification, Through the use of logie they hoped fo return English co an “undefiled” state, With untizaited zeal and imagination but limited knowledge about language change and the history of English, they added the § we now have in Island and the b of doubt and debe}; they made other changes based on equally misunderstood etymologies. They reasoned that since a subjective complement means the same thing as the noun it re~ naines, people should say Itis I rather than fe is me, that oné should use the nominative he in Bill wanted to be he but the objective him in Bill wanted Icto be him, Unfortunately, they were silent about whether to use the nominative or the objective in We discussed its being ——. By logic this should be We discdised its being my, the possessive my agrecing with its. They reasoned out rules for shall and will, should and would. ‘They legislated against ending sentences with prepositions and against splitting infinitives. These rules were based entirely on logic; no attention was given to what educated people were actually saying. For their classification of words and sentences, they folloyed the patterns set by grammars of the preceding two centuries, Fronr this tradition developed the English grammats used in schools during the nineteenth and foeenticth centuries. This traditional grammar is best known to many people in the United States from high-school text- books, college handbooks, Plain English Handbook By Walsh and Walsh, and Disrviptive English Grammar by Howse and Harmon, It followed Latin grammar in concentrating on parts of specch that are subcategorized according to case, perton, number, gender, mood, tense, etc. ‘These concepts are informative in a study of Latin, but many of the categories are hard #0 justify for English. Word order was usually ignored. Sentences were classified as simple, compound, comples, or compound-cn-aptox; elaveae were clasified as independent, noun, adjective, or adverb; phrases woe | Bnglih sland was resplled Island by analogy with French ise, alehoug the twa words sare not related etymologically. The bin doubt and debe wis acted by analogy with Lavin Gubicum and debicum, although Old French bad fst the bin these words before Enigish borrowed thersgrammars of english 5 prepositional, participial, gerund, oF infinitive, Many school grammars used the study of English grammar as nothing more than a background for a study of punctuation, subject-verb agreement, pronoun case forms, and other ‘matters of usage. * STRUCTURAL GRAMMAR During the nineteenth century, as scholars began studying and comparing large numbers of languages, many of them radically different in structure from Latin, they saw that traditional grammar was inadequate, Some missionaries and other people describing cxotic languages assiduously forced them into the framework of Latin grammar, but others realized the shortcosnings of the tradition. This dissatisfaction with traditional grammar continued into the twentieth century, when Jespersen and Poutsina found it necessary to make significant departures from traditional grammar in their monumental works on English.} Many teachers became disillusioned with the grammar they were teaching when they discovered that it would not account for many ordinary sentences that are encountered in modern newspapers and magazines; for exercises they were limited to carefully selected sentences in their texts, They often found that in their teaching they were relying inore upon observations they had made about the language than upon explanations in the texts. Discontent with traditional grammar grew to such proportions that during tne second quarter of the twentieth century a mew approach to the study of language evolved: structural linguistics. Followers of this approach felt that it was necessary to study the structure of a language as objectively 8 posible without reference to any other Ianguage, and they felt that meaning was a poor guide to the analysis of structure. Instead of talking about what a noun means (“the name of a person, place, or thing"), for example, they began looking for other devices to identily nous. Tn a sentence such as The arguments became heated, they said that the word arguments can be recognized as a noun because it has a plural ending, because it has the suffix -ment, because it follows the determiner the, and because it precedes the verb became. They attempted to analyze other grammatical clements in terms of sirtaure rather than meening, 1 Otto Jespercn, A Med Exglish Grammar on Histrea! Principles, 7 vols. (Copenhagen: Bjoar Munksgaard, 1909-1849). HA. Poutems, A GrarumarafLate Mader Engi (Ctoviegens B. Noothoff, (914-1925), Jespersen and Pouttma belong wither with the taditonal Srammarians decissed in the preceding ction nor with the steacaratsts, Glearon in has Lingsvtcs ad English Grozomar (New York, 1965) cally them “sekolacly. taal”‘6 chapter one ‘Phe structuralists based their conclusions upon analyses of sentences that chev had collected from native speakers of English, Onc of the most remarkable of the structuralists was Charles C. Fines, who obtained access xo letters written toa government agency as a cbfbus vr his analysis presented in American English Grammer (New York: Appleton-Century-Crofts, 1940). Later he obtained permission to record telephone conversations, unknown, 10 the people talking. The results of this study were published in The Siracture of Englisk (New York: Harcourt, Brace and World, 1952). Because ‘of Profestor Fries’ integrity and efficiency, access to the materials was cavefully controlled, The illustrative material in the published results of the studies had all names replaced by dashes, and sentences that could identify the speaker because of their content were carefully excluded. Professor Fries’ work was widely read and emulated by later structuralists such as Francis, Hill, and Stageberg.+ hie. work of the stnicturaiss.uromght attention. to-the-word-tingwisty fa term that had previously been used to desienate someone who studied angtages, partieu‘ly someone who spoke several different languages But after the publication of Leonard Bloomfield’s Language in. 1983, the term linguist became specialized to mean only the person concerned with the new scientific study of language. The word finguistics became popular as the name of his diseipline, Until the late 1950's, the terms strcturaist and linguist were practically synonymous in America, The structuralists made ‘many praiseworthy contributions to the study of language. ‘They challenged current attitudes and began the scholarly study of language that has apidly increased during the last «wo decades, TRANSFORMATIONAL GRAMMAR Starting formally in 1957 with the publication of Noam Chorrsky’s- Suntaciz Structures, a nest approach to the sturly of language was inaugurated. -Lnis newer gramamar has gone under various names: generative, franifermational, generativetransformatzonal, and transformational-generative, This is the kind of grammar presented in this book, and the term transformational is used for consistency and brevity. Scholars have been so impressed by the importance and potential of transformational grammar that since 1957 the majority of published studies of English syntes have wed this approach. During the 4 Avchibald A. Hil rodtion to Linge Sects (New York: Harcourt, Brace and World, 1958), W. Nelon Francie, The Stactoe of Americas Englch (New York: The Ronald Press Company, 1954). Noritan ©, Stagsberg, te Inuradetry English Grannar (New York: Hel, Rinchset and Winston, ne, 1955).grammars of english 7 past ten years there have been many developments in transformational srampar, and there will certainly be more in the future. Almost every spect of language is sill being examined. There is so much controversy among linguists that many people usc the plural in speaking of tnenapor, ational granmars; there is no single transformational grammar which fy accepted by all scholars in the Feld. Nevertheless, there are inany ideas which most tansformationalists do accept. The transformational grammarian is not content with describing what be finds in a corpus of sentences collected from native speakers, He rae shar bis grammar should enable one to produce ll hesentences ofa language, and he is as interested in possible sentences as he is in the ones actually, [ieerded. Since the number of possible sentences in English or any othey Jangwage is infinite, no one could have heard all of them. Yet native {heakers of English understand new sentences such as F spilled milk in the bsehtub and He left his shoe polish in the refrigerator. Every day the ative speaker hears, reads, and creates new sentences, sentences tick Scem so ordinary that he is not aware that they have never been used before An adequate grammar of English should enable a person to produce noe ist those sentences that have been said in the past, but all the sentences {hat & native speaker is capable of creating or understanding. Tu addition, {he grammar shoul not generate sentences that a native speaker would eject such as sThe man horrified the doer or #Boy on the roof is Notice the asterisk in the last two sentences. ‘This mark is used to indicate that a sentence or a part ofa sentence is ungrammatical; that is, no ative speaker of the language would intentionally use it. By granmaticat whe are nor celerting to standard and nonstandard usage. He ain't going is grammatical for come people (there are native spealeers who say it), Dut i is ponstandacd (educated speakers do not normaly use it). «He net ls going senck Stammatical, since no native speakers use this construction. Calling sentences like He ain't going and Irregardless of what you think, be taken te grammatical does not mean that the linguist is encouraging anyone to wee {hem when applying for a job, He is merely restricting the micaning of the Ein Sanmatical to steuctures that are-normually used by native speakers of Faglish, He efers to acceptability among various social groups or in various contexts as usage. The transformationalist is more concerned with the system that underlies the language than he is with the actual speech of an individual at any given time, All speakers occasionally stamimer, make false starts, vse tense (M) (have + en} (be + ing) Regardless of meaning, classify the following modals as present: may, can, will, shall, must; these as past: might, could, would, should,the auxiliary 2 For most verbs all combinations of the auxiliary are posible. ‘There are some nonaetion verbs, however, chat do not exist with be + ing: seem, hear, know, hace, understand, etc. (#She was seeming frightened), The copula be as in He was a nuisance docs not take the ausiliary be + ing, although thereis another verb be (He was being 2 nuisance) which expresses action. This feature will be weated asa peculiarity of certain verbs and will be included in their dictionary entries. We do not need to change mule PA for these exceptions Here is a tree for the sentence He could have been running. It contains the maximum expansion of the ausilary You have probably wondered about our treatment of tense in some sentences. 7 We have said that He his seen us con: 7 tains present tense, but He had seen us ZL Ds | contains pat tense. Its obvious that both ve piety sentences describe action that occurred he past chubaie ch thing. nh duving past tine. Alto we have called the tense in He will go present and He would go pas, although they both cleatly Indicate future time. Tn act, we have not even mentioned s future tense ‘Traditional grammar lie thece basic tenses. It defines present tense 4s an expresion of action oceurring atthe present time, past tense az action cceurring during the past, and future tense as action occuring daring future time. These terses are represenled by the fllowing forme of the verb: I see (peosent), Isaw (pas), shall see (future). In addition, there are three perfect tenses, indicating completed action and formed by the ac of have + en to the auxiliary. Many traditional grammars Hop with thee six tenses; but others add three progressive tenses, which indicate continuing action and are formed by the addition of be+ ing to the auailiay, Trad tional grammars generally assume that there iu a atice correspondence between time and the form of the verb in English Such a correspondence does not exist in actual usage, Present time may be represented by the preent-ense form of the verb as in Now | understand or Here comes Sally but iis usually represented by the present progressive form: He is sleeping; They are leaving now; !atn eating, For politeness we sometimes use “future tense” for a present action, at check: fou clerk may say, “That will be $6.50, please.” The normal meaning of the present-erse form is not present time, but rather aetion that is habitual or universally true (i.e, past, present, and fature time): Tim sa policeman; Water freezes at thiety-twa degrees; leat lunch a one o'clock every For many verbs itis imposible to state an action occurring specialty in the present time with the simple present lense: sl eat the cake now (cl 1 cat eake often andl lam eating the eake now), Yeue chapter three Past time is normally expressed by past tense: John ran to town; He had already finished when I came in; In those days | ate a big dinner. But sometimes present tense is used to express past action, as | see that he has been defeated, or When Benjy hears the golfer cry, “Here, caddie.” he begins crying, or Europe enters a renaissance during the twelfth century (the historical present). The present perfect tense reflects past time in spite of its name: He has gone; We have finished Future time is as often expressed by other means as it is by will and shall; in fact, it may be expressed by any modal or by any tense except past, with adverbials such as tomorrow indicating the time, We leave for New York tomorrow. (Present) If she stays, we must leave, (Present) ‘Ask her if she will stay. (Future) ‘Asie her if she is going to stay. (Present progressive) He has his chance tonight. (Present) He is leaving soon. (Present progress When she comes, we'll leave. (Present; Future) He Is about to go. (Present) She is to accompany him. (Present) 10. What would you do if you had a flat tonight? (Past) No one denies the existence of past, present, and future time. We do say that the forms of verbs and ausiliaries in English are not so consistently related to these thee times as are the forms of the Latin veeb. Time in English is often expressed by other means than the tense of the verb. By {form dese are only two tenses in English: present and past. Remember that by tease we mean the form of the frst auxiliary that follows the symbol tense, If there is no such auxiliary, then tense is shown by the form of the main verb. ‘Tense is a concept of form, not meaning. * Elaborate niles for the use of shall, will, should, and would have been devised by earlier textbook writers. The rules first appeared in the seven- teenth century and have been perpetuated with embellishments tothe present day, Many writers of high-school English texts and of colleye-handbooks are notorious for merely copying rules that have become traditional and never questioning their validity. Most of these texts say that shall and should ate normally used for frst pervon, will and would for second and thied persons, To show determination, one is to reverse this order, using will and. ‘would in first person, shall and should in second and Chird. Then there are many other rules that vary from text to text, such as the one requiring shall in the question Shall you be here tomorrow? since the anticipated answer is 1 shallthe auxiliary 25 ‘These rules offer a good example of the utter futility of trying to legislate English usage. In spite of assiduous drill by schoolmarms for two centuries, the mules have never been followed by many educated people. ‘Most of the better books on usage today have adopted the more sensible approach of describing what educated people actually, say, rather than trying to prescribe rules that seem “logical” or “in good taste” to someone while he is thinking about language in his study, Ie is true that in spoken British English shall and shan’t do oecur often with first person pronoun In spoken American English, however, shan’t is almost never heard, and shall appears formal, emphatic, or even affected. No doubt: the main reason for the infrequent use of shall in spoken ‘American English is that this modal did not become widely used in London English until after most of the English speaking settlers of what is now the United States had emigrated. The later rise in popularity of shall in England did not spread to this country. Also, in American English both shall and will in speech are contracted in sentences that are not given particular emphasis. T'll go is the normal spoken form, with | will go or | shall go showing either formality or emphasis. In fact, shall because ofits infrequent occurrence stands out in spoken American English and gives emphasis or determination for all persons, as McArthur’s “I shall return’ clearly indicates. The only unemphatic use of shall in spoken American English is jn questions like Where shall we sit? and Shall we dance? But even these are now being replaced by such forms as Where would you like to sit? and Would you like to dance? Except for the emphatic form, shall is rarcly found in spoken American English, In formal American English, usage is divided among educated speakers between shall and will for first person. Although | thall and we shall occur frequently in formal written English, its not at all unusual o find | will and ‘we will in articles in professional journals and in formal correspondence from. English professors in leading Amei iversities and from editors of major publishing firms, particularly when rapport with the reader is especially desirable, Should, like shall, is used differently in British and American English. In unpretentious American speech, should always means “ought to.” To most Americans, | should be glad to have you visit me either sounds affected or seems to indicate reluctance. In formal weiting, the same practice is observed. Would with first person as in | would be glad to hear from you is the usual form, exercises ‘A. Draw trees of the fallowing sentences: 1. Mabel might be upstairs, 2. Evidently he bad read the report reluctantly,6 chapter three 3. That man must hive lacked courage 4. The members would have chosen the leader then. 5. Harold might have been sitting in his room. G He will be a minister next year. 7, Yes, Murphiy has been very quiet @& The firemen wil be having trouble soon B. After each of the following formulas there are three sentenoes that may or not have the structure described by the formula. If a sentence follows notation given, write erred beside it. If ¢ does not follow this notation, wig the forraala that it does follow 1. NP + present -+M + V + NP. 1a, The student should use the library. (The car will hit the child. (3) He exa sing the song op past Mi have + en + be + place ‘a. The cow has been in the field | x. by They should have been eating. "(Oi People could have been running in the rain. INP + present + have + en + V + NP Js, We will have caught him. 7 © They ave chosen John . We are electing him 4. NP + past + M+ be + NP a. He isthe owner b, Everybody has gone homme. G They would have been conservatives SNP + present +M ¢ have + en + be + ing + V+ place a, Jobn isin the bars. & Jack wil have been running in the house. . She nvyst have been ill 6. Write the formula for each ofthe sentences below, make the addition or delet that is called for, and write the resulting sentence. Example Sally was playing. NP + past + be + ing play Add have + en: NP ++ past + have +en + be + ing + play Sally had been playing. wa Tate then. (Add be + ing) 2 “°°, fete . They drank the wine. (Add with) 7287 : 7 We had gone to the window, (Delete have 4 en) | You eould have listened to me. (Add be + ing); “They might be disappointed, (Delete may) P20", Sam was telling a lic. (Add hove + en) yoo eA ‘You could have been repairing the clock. (Delete be + ing) Las going. (Add shall) 7 1 T could see hire wel. (Delete cam) - cs ‘The waitress must be laughing at us. (Add have + en) Beesepeurechapter four LEXICAL FEATURES DIVISIONS OF A TRANSFORMATIONAL GRAMMAR In the last two chapters we devcloped the phrase-structure rules for English. ‘These rules begin with a single symbol $ and progress through ordered. expansions until no further expansions are possible. The rules are repeated. below: Pl: P2: Pa: PS: PT: Pa: S—+ (SM) Nuc : Nuc> NP + VP P3: VP Aux + MV (manner) (place), (time) (reason) Aux — tense (M) (have + en) (be + ing) roe) pee w-f[ HE) van NP -+ (Det) (PY) AP (Intens) Adj 128 chapter four By starting with rule PI and progressing in order through the rules, selecting or rejecting optional elements, we may expand § in the following way’ s i Pi Nuc i P2 NP+VP P30 NP-+ Aux+MV PL NP + tense + be + ing + MV P5 NP + present + be + ing +MV { P6 NP + present + be + ing + V +NP "P7— Det-+N + present + be + ing -+ V + NP | P7 Det +N-+ present + be + ing + V } Det +N PB Does not apply ‘There are no longer any symbols that can be rewritten since nothing in the final line occurs to the left in any of our rewrite rules, Symbols such as Det, N, or present that cannot be rewritten by the phrase-structure rules are called P-terminal symbols. A sequence, or string, of these symbols is called a P-terminal string. P-terminal string such as Det +N +} present } be fing + V + Det +N describes the structure that underlies a grammatical sentence of English, butitis not itself a sentence. It is merely a string of elements, some of which may be out of order and none of which are in their phonetic shape. The phrase structure also tells how the elements in the P-terminal string relate to each other; in other words, it specifies a structure, as exemplified by this I i Det N present be ing Vo Ost Wlexical features 29 Included in the phrase-structure section of the grammar is a lexicon or dictionary, from which we replace such elements as Nand V wi Det + N present + be + ing + V Det N ‘The + boy + present + be + ing + read + a + book ‘These, of course, are aot the only words we could have chosen for this structure; the following choices are among the many others that could have been made: 1, A socretary is mailing the letter. 2. "The painter is painting the house, We next proceed to the semantic section of the grammar to provide a meaning for the structure, The lexicon will have given meanings for the individual words. ‘The semantic section will tell us how these meanings are combined to provide a meaning for the sentence as a whole. Since very little is known about how the semantic component works, it will not be discussed further in this book. ‘The phrase-structure rules do not generate finished sentences of English. In fact, they provide only a smail number of elementary structures that underlic the sentences of English. To rearrange, delete, or adel Structures, we need transférmational rules. For example, we can provide a rule for the structure: the boy present be ing read a book that rearranges it as follows: Present be the boy ing read a book ‘Thisis the order for the question s the boy reading a book? ‘The rearranged structure is illustrated by this teee: $ Such transformational rules will be dis- ‘cussed in detail in the next two parts of | | / N Another “set of rules i necteas \\ pete oes that el us hen to Ne ronounce the string that we’ have J\ senate. Theses il be dca jn Part Four, and they will ell us how }} combinations otwords are pronouneads tebk for” example, presene 4 tat iy cece resent be the boy ing ead30 chapter four post + drop is dropped, etc. (Notice that inflectional morphemes, such fas past, en, and ing, rezain before the verbs since they are merely symbols, not phonetic representations.) After we apply these rules and those for the addition of stress and pitch, we have a sentence: Is the boy reading a book? We are describing rules that will enable us to produce the sentences of English. We cali this collection of rules a grammar. The grammars organized into three sections or components: 1. The syntactic component contains the phrasestructure and tans. formational rales and provides the structure of the sentence, 2. The semantic component operates on the P-terminal string after entries from the lexicon have been added and gives the sentence its meaning. 3, ‘The phonological component operates on the sentence after all trans. Tormations have been applied and gives the sentence its final form. I we have’enough rues in each ofthese divisions, we can generate al ofthe sentences of English, including all potential sentences that have never heen spoken belore. But the ideal grammar not only preduces all of the tersmmatical entences ofa language; it produces no ungrammatical ones ‘The grammar as we have stated ito far would produce sich wnacceptable structures as la, 2a, 34, and 4a Ja. *The smoke vanished the table. b. The smoke vanished. 2a, *An event occurred the man, b. Am event occurred, 3a, *The man brought. b, The man brought a book. 4a, “He proved. b. He proved his point. Rule PS as we have stated it rewrites MV as be plus one of several structures or as Vy) say optional NP. Sentences Ia and 2a are ungrammatical because this optional NP fas been -clected; 32 and 4a are ungeammatical because it has not been selected. TRANSITIVE AND INTRANSITIVE RESTRICTIONS ‘The lexical entry for each verb will have to state whether it permits 2 following noun phrase or not. Vanish and occur will be listed in the lexicon ‘with the feature [ NP}, and bring and prove will be listed with [ +,__NP]. The underline indicates the position filled by the verblexical features 31 before an NP. The plus indicates that the verb in question can fill the position; the minus indicates that it cannot. Some other verbs with the ‘feature [ ++__NP J are shoot, wax, and tear, allusiated in the following sentences: : 1. He shot a rabbit, 2. L waxed the car 3. Someone tore the page. ‘Some verbs with the feature [ +__ NP] at times have indefinite noun phrases deleted: 5. Lamy aie (something). 2. The suadents wrote (something). 3. He drove (something). Other verbs like these are watch, steal, and read. Whether these verbs have & noun phrase following them or not, there is no alteration in meaning, ‘These verbs sill have the feature [ + NP-] even when the noun phrase is not present, The feature merely states that a noun phrase is permitted after the verb with the meaning given. ‘One word of caution is in order. The classification ofa verb depends upon its meaning in the sentence under consideration, ‘There is a difference in meaning in He ran fast and He ran the store. We set up run in our lexicon as tro separate entrics, or as two separate words. The entry with the meaning as in He ran the store has the feature [ + NP J; the entry with the meaning He ran fast does not have this feature, but rather { ~_.NP]. This distinction may hecome complicated for verbs whose meanings are metaphorically extended. Notice fly in the following sen~ tences: ‘The bird flew out of the.room. in flew out of the room in a rage. = i an airplane. Last week T flewan airplane. ‘The dust lew everywhere, He flew off the handle We will lit fy in our lexicon six different times, one for each meaning. Similar metaphoric extensions ean be seen in rise (The sun rose; The boy ‘0se; Prices rose) and die (The man died; The car died).32 chapter four Verbs with the feature [ —___NP] may be followed by nothing or by optional adverbials, as in the following sentences 1, ‘The nurse was sleeping. ‘The nurse was sleeping soundly. 2. The boy fell ‘The boy fell to the ground, 3, She hurried. She hurried because of the rain. Of course, as we said in the preceding paragraph, different lexical entries may have the same form, Hence, hurry in She hurried is not the same word ts hurry in She hurried us off to work. Some other examples of verbs that do aot permit following noun phrases are vanish, sweat, pause, rise, and Gecur, These verbs are traditionally called intransitive verbs, as opposed: to transitive verbs, which do take following noun phrases. Many intransitive verbs can on occasion be followed by a noun that is similar im form to the Yerb: He slept a deep sleep; He dreamed a dream (ic., nouns that tradi- Uonal grammar called cognate accusatins). Since these axe apparently the only nouns that may follow thesc verbs, they do not affect the classification, ‘A complete grammar would probably introduce them by transformational rules, ‘Some intransitive verbs must he followed by adverbials if the sentence is to be grammatical: 1, We lay thers, We lay 2. He lurked outside, *He lurked. Unlike the simple intransitive verbs, these must have an adverbial of place following them, Other examples are lie (recline), step, and glance. These Verbs have the feature [+—— place]. The adverbial is not optional Sometimes wo avoid repetition we delete elements that are otherwise essential. tWe sneaked, for example, is ungrammatical by itelf. Tt may occur however, in the context How did you get past the guard? We sneaked Tinderlying We sneaked is the complete sentence We sneaked past the guard. Past the guard has been deleted to avoid repetition. This kind of deletion does not affect our classification. “There are also transitive verbs dat must bave adverbials of place following the noun phrase 1, He handed the paper to me. *He handed the paper. 2, She set che book there, She set the book 3, The doctor laid it there. #The doctor laid itlexical features 3 ‘These verbs have the feature [ + NP place J, meaning that an adverbial of place must follow the NP. With these features—{ + NP], [—_- NP], [+ place], and [ +—_NP place ]—we can account for the native speakers intuition that sentences sueh as #We persuaded and *Someone shivered the man are ungrammatical, SUBJECT-VERB_ RESTRICTIONS Earlier we said that the phrasestructure rules provide P-tcrminal symbols such as N and V and that we select items from the lexicon that are dcsignated by these symbols, If'we select just any N or V, we coud produce the following sentences: 1. The man drove a ear. 2. An eraser fell to the floor. 3. *A door sneezed. 4, He surprised my curiosity. As the Tas two examples show, we may not select just any noun of verb, but ‘we must impose certain restrictions, Notice the following sentences: 1, #The coffee prayed, 2, “Those ants talked to me, 3. *The umbrella coughed ‘These are all unacceptable because the subject noun phrases are nonhuman and the verbs require huinan subjects like the following: 1. The woman prayed. 2. Those children talked to me. 3, My uncle coughed. Tt is, therefore, selevant to our grammar whether a noun names a human for not. In the lexicon to indicate that a noun represents a human, we will34 chapter four assign the feature [+ human J; all ether nouns will be [ ~ human ]. ‘Thus, woman, boy, and aunt are [4+ human], and cat, ivy, and door are [— human}. Verbs that do not permit nonhuman subjects have the feacure [ — | — human ]__ }. The underline indicates the position of the verb, afier a noun with the feature [ — human ]. Verbs such as pray, talk, faint, worship, complain, praise, acknowledge, and thank will be listed with this feature to limit them to [ + human] subjects, Other verbs will not be marked for this feature, We can thus explain why *The door fainted and +The dog praised me are ungraramatical. In childzen's stories or other imaginative literature, of course, door and. dog may become human. Some people will argue that all ofthe verbs in the last paragraph may have animate nonhuman subjects; normally they do not make this claim for all animals, but only for pets or certain other animals of which they are particularly fond. Rather than extend the features of the verb for these people, it seems better to say that the distinctions human and nonhuman have been erased in certain cases for perhaps that dog, horse, etc, sometimes have the feature [+ human } (when they are the subjects of verbs such as worship and acknowledge) ‘The features we have given will account for the grammaticalness of sentences 1a, 2a, and 3a and for the ungrammiaticalnest of 1b, 2b, and 3b Ia, ‘The man bit me. b. «The street bit me, 2a, A child ran rapidly into the street. b. ¥A rock ran rapidly into the street. 3a, The cook tasted the soup. b, *The chair tasted the soup. However, if we restricted such verbs as bite, run, and caste to human subjects, we would not be ablé to produce the following sentences: le, The horse bit me. 2c. A dog ran rapidly into the street, 3c, The cat tasted the soup, Since these are grammatical sentences, limiting these verbs to human subjects would be too strong a restriction. Animais as well as humans can bite, run, ot taste something. We need, therefore, another classification of nouns that includes humans and animals. Dog. horse, girl, man have the feature [ 4 animate }; rose, rug, and rock have [ — animate ]. Verbe that do not permit inanimate subjéets have the feature [ — [ — animate ] ___]lexica features 4 35 ‘The need for a final restriction between subjects and verbs ean be seen in the following sentences: Ia. His coat lay on the bed. b. His curiosity lay on the bed. 2a. The cake tasted good. b. *Pettiness tasted good. 3a. An accident occurred. b. *A snake occurred. 4a. Six hours elapsed. b. *Six dogs elapsed. Sentences Ib and 2h have abstract subjects with verbs that require concrete subjects; 3b and 4b have concrete subjects with verbs that require abstract subjects. Gencrally speaking, concrete nouns name objects that can be touched or seen (cat, apple, woman, etc.). Abstract nouns do not (hap- piness, honesty, courage, etc.). Abstract néuns are not physically located in space. Concrete nouns have the feature [ + concrete ]; abstract nouns have [ — concrete ]. A verb that does not permit abstract subjects has the feature [= [concrete] __] (sit, rise, lle, ete.); one that does not permit concrete subjects has the feature [ ~ { + concrete] (occur, befall, elapse, etc.). A verh such as amaze, which may occur with fo abstract subjects, will not be marked for this feature. “The four features { — [ — human] —_],[ — [ — animate] ____], [= [~ concrete | __], and [ ~ [+ concrete] ] will be given in the lexical entries of only those verbs which are restricted as to the kinds of subjects they may follow. Other verbs, such as seem and appear, will be listed without any of these restrictions since they apparently ascur freely with concrete or abstract, animate or inanimate, human or nonhuman subjects 7 “The verb features we have given are illustrative rather than exhaustive Tor example, we could have a classification for verbs that require animate objects (surprise, astonish, cerrify, frighten, etc.). Or we might geoup those verbs together that requiee living subjects: humans, animals, and plants (grow, die, thrive, etc,). Some verbs take only abstract objects (guess, pretend, announce, evc.). A corhplete classification of this nature ‘would go far beyond the scope of this book. Some verbs are even more specific: evaporate must have a subject that is a liquid; read and weite must have objects pertaining to writing Other verbs have equally severe limitations on their subjects (disperse, dissolve, corrode, etc.) or their objects (sing, whistle, say, etc.). A complete lexicon would deseribe these idiosyncracics.chapter four Here are some verbs with their syntactic features as they would be given in a lexicon: . admire eat [+—_NP] [—[ — animate] —__] [ (-__NP] [= (+ concrete J} DETERMINER AND NOUN RESTRICTIONS ‘We have now classified nouns as human or nonhuman, as animate or inanimate, and as conerete or abstract. These features are needed in accounting for restrictions that hold between nouns and verbs, There are also restrictions on which determiners a noun may take: Ja. +1 saw bug on foor. b. I saw bugs on the floor. 2a, *Phe honesty is an admirable trait. b. Honesty is an admirable trait c. Honesty is admirable trait. 3a, *The William entered the room, b. William entered the room, To account for the differences between the grammatical and the ungeam- ratical sentences above, we muse discuss additional noun features. Nouns are either common { + common ] or proper [ — common J. ‘The traditional definition of a proper noun as the name of a particular person, place, or thing and of a common notin as any one ofa clase works in sany cases: Joe vs, boy, Omaha vs, city, etc, On the other hand, the names of scasons of the year, which are traditionally considered common nouns and are not capitalized, are as specific as nantes of the months, Which are proper nouns and capitalized. Also, the devil seems to be as specific as Satan. Normally proper nouns do not take plurals or articles, dhe word the in such names as The Hague and the Amazon hiding Takich as part of the nhame and not a determiner, But even here there are problems, No one would question calling Mr. Smith and Mrs. Smith proper nouns, but when we speak of them together as the Smiths, we have added both an article and a plural morpheme. The following sentences also present problems: Thelexical features 37 Americans landed three Sundeys ago; A Canadian has two Chevrolets; There are two Jims in my room; He Is not the John Smith who lives nexi-door to, me, ‘That there is some real system underlying the distinction common and proper can be seen in a sentence such as Blamp opened the door. ‘The native speaker recognizes Blamp as the name of a person (or pet), although. itis a now word to him. Our grammar should be able to draw similar conclusions. No doubt much of our confusion is caused by a too clore association of the noun designation proper with an initial capital letter. ‘The differences between common and proper nouns is a subject that needs further research, Some nouns name objects that can be counted, whereas others do not. ‘The sentenee i read a book can be altered without changing the structure to | read two books or three books or four books. Book has both a singular and a plural form, and various numbers can be placed in front of it. We call book a count noun and say that it has the feature { + count]. In the sentence I drank a glass of water, the word water cannot be made plural, even if we alter the number of other nouns and verbs in the sentence, nor can we add numbers directly in front of the word. Unlike glass in this sentence, water is a nancount noun and has the feacure [ ~ count ]. The following sentences provide further examples of this feature: 1, The horse bit me. [+ count] 2. Those doves flew away [+ count ] 3. The horse ate some oats. [—count] 4. The ground needs moisture. [ — count ] tis possible for a noun to be count in one sentence but noncount.in another. In the sentence The ground needs water, water exhibits the feature [~ count}. But in He has seen the waters of the world, it exhibits [ + count]. The word water doet not mean exactly the same thing in the two sentences. Notice how the following sentences vary this feature: He ate two eggs {+ count } |. He got egg on his shirt. [— count } He threw me a stone, [+ count J The house was made of stone. [ — count] I spilled a Coke. [+ count] T spilled a glass of Coke, [ — count} Noncount nouns, such as those ia the even38 chapter four often seem to mean substance or material (they are often called mass nouns), whereas the count nouns mean the individual items. A dictionary usually lists these meanings separately. We will treat the count noun stone in sentence three as a different lexical entry from the noncount stone in sentence four. Sometises a noun is noncount although we could count the items it names. Furniture, for example, in He owns much furniture, is noncount, since we cannot say one furniture, two furnitures, chree furnitures, etc, Yet we can count the pleces of furniture. Similarly, sand in He got sand in his shoe is a noncount noun, although we can count grains of sand. Some abstract nouns are noncount and do not permit determiners or plural morphemes: *The honesty is good; *They have courages, Other abstract nouns, such as idea and trait, are count nouns and take determiners and plural morphemes freely. ‘The features [ + count ] and [ — count ] determine which determiners may accompany a noun. If the noun is [ — count], 2 snay not be used: #A happiness was felt; +A furniture was in the room. If the [ — count ] noun is [ —concrete J, normally the, this, and that may not be used: ‘felt che sadness; #He wasfull ofthat enthusiasm, Nouns thatare [ — count] and [4 concrete] take the freely: The ofl is leaking on che furniture Count singular nouns may take a or the; they may not exist without any ‘determiner; A man sat in the room; *Man sat in room, Count plural nouns may be preceded by the or by no determiner; they may not be preceded by a: The students came in cars; *A students came in a cars. Features of a noun can be represented in a matrix, in which we have rows of positively or negatively specified features. Below are the features for the boldfaced nouns in the following sentences: Mary loves truth. The woman owns a cat ‘woman Mary cat rath N + + + common, + 7 + count + - + - + concrete + 4 + a animate + + + - human + + - - |A matrix such as this contains much redundant inforamtion. When we classify woman a count, concrete, animate, and human, we are giving much obvious information, since all human nouns are necessarily count, Concrete, and animate as well, By omitting all redundant information, welexical features ‘can turn our mnatrices into more usefil forms: woman Mary N . +t + common + - count concrete human + + - A complete grammar would contain rules such as one that automatically applies the features | + animate ] and [ + conercte ] on all nouns with the feature [+ human]. All features that are predictable by rules are left unspecified in the matrix, THE LEXICON ‘The lexical features that we have presented in thit chapter will be included in the lexicon, or dictionary. Although there are still many tunanswered questions regarding the form of an ideal lexicon, the kind of information it should contain is clear. First, there will be a citation form of ‘the word, expressed. in phonetic notation. Second, the idiosyncratic proper- ties of the word will be given; these will include the features we have presented and all irregularities (the plural of foot is feet, the past of sing is sang, etc.). ‘Third, the meanings or definition of the word will be given, ‘The syntactic component of our grammar will make use of the features of the word, and the phonological component will make use of the phonetic citation form jonal irregularities. ‘The semantic component will use the definitions and the features Lexical entries will look something like the following: 1, fmaus) +N a small rodent [— rule M3 + common Jmaus/ + pl — /mais/ ++ animate = human 2. fmaus} +N a timid woman (ee M3] | {onmon droaus! + pl —> /mais} + human 3. frmausy +N 2 black eye {—rule M5} + common frmaus) + pl — maisl + count + concrete = animate0 chapter four ‘The direction [ — rule MS J tells us not to apply the rule for regular noun plurals, which would yield *mouses. Current dictionaries seem to indicate the plurals for entries two and three as we have given thera. This infor mation may be inaceurate; for these uses there may be ne plural forms, or the plural may be mouses. exertises A. Use features to explain wi fhe flowing sentences are ungrasnmatical: 1, #The perseverance is a virtue. 2, Fic has read gail,» 4 ' 8, 4A dir flew into the room, 4, *Despair dropped to the Boor. 5, #The eagle prayed for an how 6. *They handed the book. = ~ 7. We vanished thespot. ‘bee tre-coughed loudly. = 9."+The bread dipped. 10. My bom elapsed 2, Give the features of the baldfaced words 1, The monkey chewed the food slowly. 2 A student coughed loudly. 3. He glanced atthe water. 4. Bob handed a ruler to me 5. The accident occurred yexterday G. Examine the folowing sentences and decide why.samg of them aré ungrammat= ica. What generalizations ean you make aboue much many, fewer, aad less? ‘Much energy was spent on this project. Many apples were in the bash Much children were in the roomy |= Fl ate Jes apples than I did He has fewer finde th Se hat ls Thave. fidence than Jane hs, 1 x 3 4. Many dandeufl was in his hai 6 Fle has fewer poise tha Tpart two TRANSFORMATIONS |chapter five THE NEGATIVE TRANSFORMATION ‘The phrase-structure rules can produce the structures underlying such sentences at Thote boys might have been swimming in the lake and The manager wrote a letter. They cannot produce such structures as the follow- ing |. The manager didn’t write a letter. Did the manager write a letier? Who wrote a letter? ‘Whar did the manager write? ‘A letter was written by the manager. |. Because the manager wrote a letter... ‘The letter that was written by the manager ‘The letter written by the manager For the manager to write a letter... 10, The manager's having written a letter All of these structures eeem to be related in some way to The manager wrote aletter. Thesame relationships are found in all of them: the manager js the one who performed the act of writing, and a letter is the result of this action, In spite of differences in form, there is a similarity in meaning in all the structures, Transformational rules are used to produce these changes in form. Earlier we listed several sentence modifiers: yes, no, etc. To these we ‘add not, which distinguishes a sentence such as John could sing well from the negative sentence John could not sing well. By sclecting the SM not, we can derive a structure as shown on page 44. This gives not John past can sing well, which is not grammatical. It would be grammatical if ‘we changed the word order fo John past can not sing well (John could not achapter five SH el ao Se _ siructure, A strustare goneraied oS Fy oN Fhraseiracture and lenical rules yeh enct John pas can sing well inden ve Ip structure, A deep structure that fey AN ee Englisheentence such as fob eouid cy sing well is cafed a suruee sructas 4 Ailgrunmatcal Englah sencness a suchce structures; underiying each ccc in deep sruetute, In Ine clapton we will modify this statement somewhat. Bath despa vanes snes tre more absitact than we are preening them her, since’ passage Tules have nat yet heen applied. However at this sage itiveater ong the transformational rales ewe ute a form auch as not Toke could sink ne ; than the abstraction that dhe actual deep structure, All native apeaker of Bnglah are capable of producing an Infaie number of surace structure, of grammatical Englah setiaens Ge Grrpous unto to yu leg was esata eee aay ee five-year old can do that perfectly. Our main purpose in studying these structure i to learn to make accurate obseratoer shout hog: Engh operat, By rgetve we ave telering to aestenee hegaiion, wot eee negation. That isnot probable has the SM not, That ie improbaole dee i not have this SM and i therefore ot pat ofthis inunediatestudy althongh the procees we are employing in thi chapter could caly be eerdcd account for such negative formas an imprebebe, Ty mincing the optional Sb not, we un generate x miiber of deep seructutes lke thse onthe left below ot John past edn sing wall 1, not Jerry could hear me Jerry could aot hear me, 2. not Bill has received it Bill has not veceived it 3. not they are going with us They are not going with us, ‘ 4, not they have been doing it They have not been doing it We need to formulate a rule to transform the deep structures on the left to i the surface structures on the right. At first glance you might say something like “Moye not to the position after the auxiliary.” In the first sentence, not Jerry could hear me, thir would work, but the third sentence would become *They are ing not go with us (remember that the ing which accompanies the auxiliary be is part of the auxiliary). We must refine our observations. In the surface structure, the negative particle not follows pactthe negative transformation of the auxiliary, but not all of it. In fact, it follows only the first auxiliary after tense. In not Jerry could hear me, could is a case of past + can. Can is the first occurring auxiliary; therefore, not follows it in the surface structure. We use the abbreviation Ayx! for the first auxiliary that comes after tense, Tn the following sentences Aux! has been marked: 1. they past can be ing go 2. we present have en eat 3. those dress Pl present must have en be red So long as the auxiliary contains something besides tense (a modal, have, or be), the first clement following tense is Aus! Our transformational rule for the correct placement of not should read something like this: “Move not to the position following the first auxiliary after tense.”” Since we will have a number of transformational rules, it will be advantageous to write them in a conventional abbreviated form, ‘The rule can be stated as follows: not +X 4 tense + Aux! | Yop Xf tense ++ Aux! + not + ¥ ‘The information on the left ofthe arrow describes the structure to which the rule is applicable: one with the SM not and an Aue, If either of those conditions is not met, the rule does not apply. ‘The information on the right of the arrow describes the structure alter the change has been made. The double arrow means that this is a transformational rule rather than a Phrase-structure rule. Whereas phrase-structure rules merely expand elements, such as Nuc into NP and VP, transformational rules rearrange, delete, add, o substitute elements, thereby altering the underlying structure of the sentence, ‘The symbol X stands for anything coming between not and tense. such as ahother sentence moder oF nou PHTEE- Since the rule operates the same way regardless of what follows not, we can simplify our rule by using the symbol X for any structure coming between not and tense. Similarly, Y stands for anything following Aux’. This may be ether aux. tiaries, a’verb, and anythuig-that follows a verb Since the same process applies regardless of what follows Aux?, we ean improve the rule by using the symbol ¥ for this For the deep structure not they present can hear you, we can illustrate the rule in the following way: not|X tense Aust) =X |tense | Aux!jnot|¥ not|they|present) can |hear you they|present|ecan fot |hear you ‘This gives They can not hear you, after the phonological rules have been.46 chapter five applied. This process can be illustrated with trees. Here is the deep struct (> Le rit \ rot they present con hear you ‘The negative-transformation rearranges the tree to provide the following 4 surface structure: q |} JV A | « ‘ney present edn not her you [After the application of phonological rules, we have the sentence They can not hear you. Before going further, you should practice with the following structures. Remerber that you are trying to understand the process and the anechanics Hy of the rule, not those apples were smelling rotten not Estelle would have done that not you are reading fast enough ! not Lucille will have finished by chen not we had heard the newsthe negative transformation : “7 Some sentences have only tense in the auxiliary and, therefore, no ‘Aux, For these sentences tie rule we have given will not apply. We need to write new rules for these sentence; ten we will see how we can modify the rule that we have aleady formulated, Examine the following deep ttructures onthe left and ther corresponding surface structures on the right 7 Wo not they are our fronds They are not our fends a not Jaws indy ne was ot end es a“ not the bird was there ‘The bird was not there. Like the sentences with Aux}, these demand a rearrangement of the structure ‘This time not is placed after be instead of after an auxiliary. Notice that in these sentences be is not an auxiliary, since there is no verb following it and since there is no ing on the next word, The be in these sentences is part of the MV. We write this rule as follows not + X-+ tense + be + Ys X-+} tense + be 4 not + ¥ Sri ttt tliat tp arte a ~ D> a swe ZN ak Wy \ rense 0é Be 43} no! fhe awer past be true The rule transforms the deep structure into the following surface structure: tense bg not vf | | # b the newer pdst be not tele The answer was not true,chapter five We have now covered those sentences with some element in the auxiliary in addition to tense; of the sentences with no stich auxiliary, we| have covered those that contain be as part ofthe MV. That leaves only those sentences with only tense in the auxiliary and with verbs other than be in the MV. The following structures illustrate the tsansformation involving ‘these verbs: not we play often We do not play often not they taste the salt They do not taste the salt not Terry catsearly Terry docs not eat eatly. not the janitor did it ‘The janitor did not do ie not the man sees me ‘The man does not see mae, In the surface structure not comes before the verb and after tense, which is attached to do, If we omit do from the surface structure, we obtain the following: We present not play often. They present not taste the salt ‘Verry present not eat carly. ‘The janitor past not do it. ‘The man present not see me. ‘These are not grammatical sentences, since net cannot be altered to show a contrast between past and present. To provide a grammatical sentence, we add the word do. Although this word has no lexical meaning, it can carry the tense morpheme: do and does in contrast to did, These sentences require two rules: not + X+ tense + V+ Ye X 4 tense + noe + V. X ftense +A + YoX+ tome + dot A+ ‘The negative rule moves not between tense and the verb, In the second rule, A stands for any morpheme other than an auxiliary or a verb, Any time tense appears before any such morpheme, we add the word do ta carry the tense. To coavert the deep structure nat we present jump here into a surface structure, we apply the negative and do transformations as shown in the following trees.the negative transformation not we peetent jump here ‘The negative transformation applies to this deep structure to produce the following intermediate structure “1 wey | | present not jump here Ne | u, Bue ‘Now the do transformation applies to produce a surface steucture: we present do jump — here After the application of phonological rules this becames We da not jump here.chapter five We have now given three rules for the negative transformation in English, as follows: not ++ X 4 tense ++ Aux! + Y=>X 4 tense + Aux! 4 not + ¥ not +X + tense + be +V3X4 tense fbe 4 not } not + Xf tense-+V + Y=sX4 tense-+ aot +V +7 ‘There is a great deal of repetition among these rules. We have a means of combining the three: Aux! ] Aux! 4 not not +X tome] be | ¥>X+eense| betnot | ¥ v not $V ‘The square brackets indicate that corresponding items must be selected on both sides of the arrow. If Aux1is selected on the left, then Aux? + not must bbe selected on the right; if be is selected on the Ieft, then be +. not must be selected on the right; if V is selected on the left, then not + ¥ must be selected on the right. exercises ‘Transform the following deep structures into surface structures by applying the negative transformation and, where applicable, the d transformation not John preseat be in the room not of course the children past eam go with us not those chairs present need paint now not she present look tired not they present be the leaders ‘not the boys preseat will have arrived by then not no that cat present rezemble my sister's cat - not they present have seen me here not Patsy past be friendly 10, not his foot past become infected Study the following sentences and determine what governs the use of some and any 1, Tread some new magazines. #1 read any new magazines 2. Some people were at the doce. *Any people were at the door, 3. Teisn't in any of the drawers, #It isn’t in some of the drawers 4, He dida't have any money. #He dida’t have some money. 5. - Some dogs were in the yard. *Any dogs were in the yard. Write five original deep structures that contain the SM not. Re sure that you 1p account for all of them with the phrase-structure rules, and use variation in dheselection of auxiliary elements. ‘Then transform them into surface structures.chapter six QUESTION TRANSFORMATIONS ‘Transformation is the process that converts deep structures into surface structures, ‘The negative transformation involves a rearrangement of structure, as when we move not to the position after the first occurring ausiliary or after be. In the case of sentences containing only tense in the auxiliary and a verb other than be, the word not is shifted to the position aficr tense, and then another transformation adds do. This second trans- formation involves addition, rather than reacrangement of structure. In addition to these two procestes, a transformation may delete material from the deep structure: He can’t sing well, but I can sing well. He can’t sing well, but I can, (sing well is deleted) (Or it may substitute something for a morpheme in the deep structure: When Jane saw me, Jane screamed. When Jane saw me, she screamed. Some transformations involve a combination of these four processes: rearrangement, substitution, deletion, and addition. Phrase-structure rules do not perform any of these processes; rather, they expand elements (NP, ‘Aux, etc) into their constituents English has two main kinds of questions: those that are answered yes or no (Are you ready? Did he leave? Should ! stop?) and those that arc answered by other words (Where are you going? Whose book are you reading! What is her name?). It will become obvious as we go along that these two Kinds of questions are formed differently. We eall the first type syeslan questions, the second type WH questions, since many of them begin with ‘words with the first letters wh, In addition, there are questions such as Sue Js going, isn’t she? and Sue isn’e going, is she? Then there is the’ echo sl82 chapter six question: You heard him come in? We will not be treating these kinds, quesnons, although dicy can easily be handled by the same processes use for yes/no and WH questions. Iewould be possible to derive yes/no questions from related declarati such as the following: 1, Tom is sick. Is Torn sick? 2. They have already left. Have they already left? 3. He heard us Did he hear us?, 6 With this approach we would have the same morphemes in both structure (except for do in Did he hear us?), but the transformation would chang the meaning of the sentence. Is Tom sick? docs not mean the same thing = Tom is sick. A principle of our grammar is that transformations affect th form of a structure but not the meaning. By means of the negative tran formation we derived He will not go from not he will go, both of whic have the samc meaning. He will not go is not derived from He will go, bu is merely similar to it, Likewise, Tom Is sick cannot be the deep structure for Is Tom sie although the two arc similar. Just as the idea of negation must be present i the deep structure of a negative sentence, so the idea of interrogation must b present in the doep structure of a question. This idea is expressed by th SM Q, which indicates that the structure is 2 question. More specificalh, it may be interpreted as meaning “I request that you answer yes or no to te yuestion. .. 2” The sentences on the left below axe deep structures that have thy sentence modifier Q; those on the right are surface structures: 1. Q she could sing well, Could she sing well? 2. Qthe book has become wet Has the book become wet? 3. Q the bell is ringing now Is the bell ringing now? ‘The process of forming. questions, like that of making negatives, is not new to you. When you were very young; you incorporated into the gramme Phat you were leaming certain rules for forming questions and negatives ‘These rules are stil part of your grammar, but you are probably not conscious of the intricacies of this grammar, and your observations abou! it may be inaccurate. You might say that you are moving the verb in froat of the subject noun phrase o form a question. This process, of course, would give *Could sing she well? for the frst question above, and you would begin refining your observation, as you did in describing the negative transformation. Tense and the frst avsiliary (tense + Aux!) have, been‘question trafsformations 33 placed in front of the noun phrase in the surface structure; Q has been Aeleted. ‘This transformation, like the negative, involves a rearraagement of elements, This part of the yes/no rule can be written as follows: QA NP + tense } Aux + Xe> tense + Aut + NP This rule will apply to a doep structure lke this: @ the belt pretent be ing ring nb It produces the following surface structure: ls the bell ringing now? Now we will examine sentences with no Auxt: 1, Q.the men are lucky Are the mea lucky? 2. Quhe was our supervisor Was he our supervisor? 3. Q Betty is at home 1s Betty at home? When tense is the only clement of the auxiliary and the main verb is be, the subject noun phrase changes position with tense and be, and Q is deleted: Q4 NP + tense + be 4 X= tense + be +} NP$7 54 chapter six so far we have seen a parallel between this transformation and the iary other than tense behave similarly to negative. Sentences with an aun a eervnees with no sich auxiliary but with be as the sain verb, Sentencss Teh only tenae as the ausiliary and with verbs other than be behave Wicrontty ander the negative transformation, Let ws see whether this ! patalicl i oxtended to the yes/no transformation: 1. Q.John read my letter Did John read may letter? 2, Q the teachers cat here Do the teachers eat here? 3, Quthe knows my name Does she know my name? If-we omit the meavingless word do from the surface structures, we arc let se amuctures sich as past John read my letter. When there is no etinry other than tense ancl the inain verb is not be, tense and the now, ! Rhvaxe change places, 30 that Q she present know my same becomes Prrsant she know my name. Since present cannot be attached to she, we Freiy the do insertion rule: present do she know my name, Thess sch: can be shown as follows ia oe Bee present know oy ee ° By the yes)no transformation this becomes:question transformations ton appli: , yee «1 Not has been included in parentheses to allow for such sentences as Isn't she going with us? ‘According to our rules, we can have a deep structure like Q not she is going with us. We apply the transformations in the order (1) negative, (2) yesjno, (3) do. By negative we get the intermediate structure Q she Isn'e going with us. A structure like this that has had one or more’ trans formations applied to it but which still is not a surface structure is called an intermediate structure, Then the yes/no transformation applies to produce the structure Isn't she going with us? Since tense can be attached to be, the do transformation is inapplicable. Now let us examine the other kind of question, the WH question, as in What is he saying? This surface structure is derived from Q he is saying something, or proforably Q he Is saying NP-WHI After the application of the yes/no transformation, we have the intermediate structure Is he saying NP-WH? ‘The WHT transformation substitutes the interrogative what for the noun phrase and shifts it to the beginaing of the sentence: What is he saying? These processes are illustrated by the following trees. Here is the deep56 chapter six ‘The yes{no transformation applies pretest bene tg say WH “The WH transformation then shifts the NP with WH atiached to it to beginning of the sentence and substitutes what i, Neo Aux NP | we Ne mY | | , tot prefotobe he itg ty Notice the process in the following sentences. The ones on the Tet ster line sruetaret that have undergone the yes/no ttansformatia| eee he right have bad the WH transformation applied to them 1. are you reading NP-WH ‘What are you reading? 2. has she torn NP-WH ‘What has she torn? 93, were you giving it to NP-WH — Who(m) were you giving it to? 'A noun phrase that has WH attached to it is replaced by what or whe ‘moved to the beginning of the sentence Teuhe noun phrase with WHL-attachment is the object of a preposi her dhe whole prepositional phrase or just the noun phrase may be mov Thc chetes is enticely stylistic, What are you writing with? being less for thas With what are you writing? We can express this transformation 1 sway: x 4 (beep) NAHE + => (Prep) (YRCe] +question transformations 7 In the structure are you reading NP-WH now, X is everything before the noun phrase (are you reading) and Y is everything after it (now). Either X° or Y-may be nothing, as in Are you reading NP-WH, ¥ is nothing As we have stated the rule, there is no way of knowing whether Q you saw NP-WH will become What did you see? or Whom did you see? ‘This choice between who and what depends on whether the NP has the feature (+ human ] or not. ‘The deep structure will attach to the NP not just WH, but also [+ human] or [human]. 'The lexicon will give who as [+ human j and what as { — human } But noun phirases are not the only structures that may have WH attachment. We may find it on an adverbial of place, as in Where are you going? which derives from are you going Adv-p-WH. Various adverbials i may have WH attachment; the interrogatives that replace them are listed below, along with a possible answer: Adbverbiat Replacement —Ansteer place (Adv-p) where theres in the yard time (Adv-1) when then; tomorrow manner (Adv-m) how carefully; with care reason (Adv-r) why. because of her; for me In addition to noun phrases and adverbials, determiners may have WH atachment. Whose replaces 2 possessive (my, his, John’s, ctc.); which (and sometimes what) replaces possessives or demonstratives. The WH transformational rule may be summarized as follows: ree es Dewitt { wc | (Prep) NP-WH ee) [e MW exercises ‘A. ‘Transform the following deep structures into surface structures: 4. Q.you present be going to the movie 2. Qithe student past see the fag, $ QT presenr may leave 4. Q.you present be afraid 5. Qithey past knovr yout58 chapter six B. Give the deep structure from whieh each of the following sentences was trans. formed: ‘Are you watching the clouds? Have you eaten today? Does she know his address? Were they in the drawer? Dia Bill plant the tree? 6. Our nile for the yes/no transformation changes the deep structure Q you present faves dime to Do you have a dime? Tt will not produec Have you a dime! Similarly, the negative transforination of not you present have a dime produces You don't havea dime, but not You haven't dime. Have you a dime? and You haven't a dime aze still common in British English, although they are dying out de fmerican English. Can you revise our rule or add a mew one to take care of these structures? 1D. Perform the WH and do transformations, where applicable, on the following interanediate structures: |. were they singing NP-WH 1 was she answering Ad have they anawored Det-WBi questions past you see NP-WEL ‘Transform the following deep structures into suface structures, performing the wis in the ones (I) negative, (2) yeajno, (3) WH, (4) do, and only one 11 Q Bok will epenk Ady WH D Got he i going Ader WE, 5 Qhe ote with NP-AVE 4 Gor you have found Det WEE book 5. Qahe wanted NP-WE 1B Ging ahe deny srocture fromm which ths seteice was desi: What could the rman have boon 60g? Fyn the dnp structure give the word hat are represented by the owing: unl, Vo M, be, Aus, MV, VP. 2 Which ane ie oad in ha ventence? Which word shes dhe tense? ee Setence why Ib the ng form of do tied instead of dos oF don? Miho ts the on fom of be mecesary? (Anover these questions fom the standpoint of form, not meaning.) 44, Periorm the yes/no transformation on your deep structure, From the rewtite fle for che WH traraformation, which words in this intermediate structure axe represented by X? by ¥ G. Why do modals not have en or ing forms? HE In the yesjao rule for sentences containing auxiliaries, why do we specify Aux! instead of just Aux? 1. ‘The verb have has an tng form (having), but the auxiliary have does not. Why?chapter seven TRANSFORMATIONAL PROCESSES We have seen that the phrase structure section ofa transformational grammar enables us to produce a limited number of simple structures such as the following: oN We call this an underlying or deep structure. Since this structure contains the SM , it meets the conditions for the yes/no transformation, which Fearranges the structure