CHAPTER Language, Linguistics, and Literary Analysis O, all the aptitudes and behaviors which characterize human beings, language is the most uniquely human, and q possibly the most important{) It is around us everywhere, in speech, writing, sign language, or simply in our minds,as we dream, remember a conversation, or quietly think out a problem, It is a vehicle of power, a means by which we control, create, and reserve “The question “WHaUTS language?” has been asked from remotest times, yet its answer is still fac from clear. The more we discover, the more mysterious and complex language appears to be. One thing seems certain, however: Language is a capacity that distinguishes human beings from other creatures. Many myths bear witness to this. For example, according to the Mayan sacred book, the Papel ‘Vuh, after the Creators had made the earth, carved it with mountains, valleys, and rivers, and covered it with vegetation, they created the animals who would be guardians of the plant world and who would also provide food for the gods and praise their name: Abd the creation of all the four-footed animals and the birds being finished, they were told by the Creator. and the Maker and the Forefathers: “Speak, cry, warble, call, speak each one according to your variety, each, -according to your kind.” So was it said to the deer, the birds, pumas, jaguars, and serpents. “Speak, then, our names, praise us, your mother, your father, Invoke then, Huracan, Chipi-Caculhd, Raxa-Caculhé, the Heart of Heaven, the Heart of Earth, the Creator, the Maker, the Forefathers; speak, invoke us, adore us,” they were told.2 LANGUAGE, LINGUISTICS, AND LITERARY ANALYSIS ¢ men; they only hissed and ¢ them speak lik HL m0 Re ee co make words, and each screamed But they coul nable to mal screamed and cackled; they were in a different way. | | , . When the Creator and the Maker saw that it was impossible qh to talk to each other, they said: “It is impossible for them to say our names the names of us, their Creators and Makers. This is not well,” said the Forefathers to each other." for them Asa punishment, the birds and animals were condemned to be eaten and sacrificed by others, and the Creators set out to make another creature who would be able to call upon them and speak their praises. This creature would be man. ‘As the Mayan myth suggests, many animals have rudime nication codes, but in no case does the communicative capacity of other animals even remotely approach our own. Efforts to teach sign language to our closest relative, the chimpanzee, have had only moderate success. Moreover, language is different from all other forms of human behavior. We speak of “body language,” “computer language,” and the “language” of music and painting, but in all these cases, “language” is used in a metaphorical ‘or extended sense, extended from “human language” to fields of human endeavor and experience where certain characteristics of language can be found. None of these senses, however, combines all the aspects of language— its symbolic nature, its systematic internal structure, its creative potential, its ability to refer to abstractions or imaginary objects, and its ability fo be used in talking about itself—and none plays the central role in human affairs that i ae All but the most rudimentary forms of_social organization depend_on language, as do all but the most rudimentary technological achievements. Jt is chiefly through language that human communities c 1 and change their siractureS, a institutions whict eiiBody community aspirations and shape comimunity fe. Without language. the astomaleen or gE and customs which we call culture would b ie of Thus, when you use your knowledge of language t ° Hmpossible, Popol Yuh, you become a new link in a chain of human Cone eon te stretching back into the distant past, a chain that ont ‘nee ponnunisstion The power of | matically attested to invone ch nee ! stories of\Judeo-Chiistian mijthology \the sto 7 to in one of the oidest Book of Genesis, According to this story, cores see ote Flood, the descendants of Noah decided to settle arene ater the Great land of Shinar.” There, instead of setting upa sociery ore 8 Plain in the God, they challenged God's authority By dona Subordinate to the will of would reach heaven. Recognizing the implicants genie & tower which regained control th ie a mPlications of this polled rough a linguistic stratagem, He defiance, God peak different languages, so they could no longer undone y ne People to Then he scattered them across the face of the cat ertand one another. ntary commu- el in theLANGUAGE, LINGUISTICS, AND LITERARY ANALYSIS 3 The story of the Tower of Babel is no longer used to explain the diversity ‘of languages in the world, but it does demonstrate rather dramatically where | the power of language lies. As God saw very clearly in the story, the people's ability TO build their tower rested on their capacity to communicate with one another, to agree on a plan, to delegate and” coordinate labor, and to give order abled the people to pool their energics in a construction ‘ten the power of God himself, To meet this challenge, God destroyed not the tower itself, but the linguistic unity which | made_the ffort_possiblé. A” wise move indeed, for as long as | the unity was there, the tower could always be rebuilt. In fact, God concluded | } that as long as the people had one language, “ nothing will be restrained Tre them, whic They have imagined To do. SS sears re tinporiance Tiniuage has in our lives, we tend to take little notice of it, wt least not until something goes wrong, as when a person grows up deaf. or experiences speech Toss of one kind or another, or simply tries to communicate in a foreign country. We think of language, too, when it becomes a political-issuc, as in Belgium or Quebec or the southwestern United States, or when it becomes a cause of bloodshed. And we think of it when we become involved in using it to create special effects, as in poetry, or in the construction of particularly difficult messages, like letters containing bad news. Even when language has been studied in the past, it has been studied mainiy as an adjunct to other disciplines, such as philosophy, psychology, religion, literary criticism, or the art of persuasion: Only over the past 150 years or so has language been studied consistently as a discipline in its own right. During this period, considerable advances have been made toward an. understanding of the internal structure of linguistic systems, that is, of the processes by which smaller components like words are combined to form utterances. Not as well understood is the way language functions in the lives of its speakers. This area, language usc, has come to the forefront only in the past ten or fifteen years. ___ This book introduces most of the major questions a linguist asks when investigating the structure and use of contemporary English, and shows how the linguist’s observations can be brought to bear on the study of literature and the phenomenon called “style.” Many of the questions the linguist asks can be considered as parts of a single question: “What does it mean to ‘know English’?” The linguist who poses this question is chiefly interested in making explicit the internalized knowledge which English speakers possess that enables them to speak and understand the English language. Another important arca of linguistic investigation can be considered in the questio How did English come to be this way?” This question, a historical one, is much too large to be introduced thoroughly in a book of this kind, although we will be frequently touching upon it in passing.” , Before turning to the structure of English, we will devote the next fewLANGUAGE, LINGUISTICS, AND LITERARY ANALYSIS ( discussing some of, and of the linguist’s app’ of human language in pages to WHAT IS LANGUAGE? Symbols We have said that language is symbolic. What exactly does this mean? First of all, it means that Ianguage involves signs, that is, entities which represent or stand for other entities the way a plus sign (+), for instance, stands for a certain mathematical operation, or the way a black armband is a symbol of mourning. ‘in language, the signs are sequences of sounds, though these cur: 6e transferred into visual signs, as in writing or the gestural sign language of the deaf. According to one theory, the relationship between an object, whether real or imaginary, and the sign which stands for it can be of three types.? If the two are associated by a physical resemblance, like an object and a photo- graph of it, the sign is called dan icon lif the relation is one of physical proximity, __ as between smoke and fire, thunder and lightning, spots and measles, then the ‘sign (smoke, thunder, spots) is called an index. If the relationship is one of ‘convention, that is, one which has to be learned as part of the culture, like the relation beween a black armband and mourning, then the sign is called a symbol. Of course, these classes are not completely distinct. Some culturally learned information is needed to interpret an icon like on the door of a toilet, or beside a road. The three types of signs are basically on a continuum, with a sign like a skull and crossbones toward the iconic end, and 2? = $ 3 x ,or-a national flag toward the © symbolic end. : Language is mainly symbolic. in that the relations between the sound sequences and their meanings are conventional and have to be learned. There is no natural connection between the sound sequence f-o-r-k and the pronged utensil we eat with; we could just as easily use the sequence k-r-o-f as the sign for this object, or the sequence t-c-n-¢-d-o-r, as in Spanish. Some words can be thought of as iconic or indexical, but even these are at least in part conventional. For example, [Link] we call “ on ic” iconic, in that ns jome way resembles what tI fs 10. One obvious example is cockadoodledoo. But if this word were truly iconic, we would expect it to be the same or virtually the same in all languages; yet even in languages as closely related to English as French and German there are seniicant differences in the consonants and vowels and even the number of syllables: French cocorico, German kikiriki. In Japanese it is kokekoko.WHAT IS LANGUAGE? 5 Similarly, ouch is in part an indexical sign for pain, in that it is normally uttered only in response to pain. But if it were completely indexical, that is, if its relation to pain were as natural as the relation of smoke to fire, we might expect the identical form to occur in many languages. But it does not. For instance, the French expletive with the same meaning is aie (pronounced like English 1). tn Japanese it is alta, in Russian okh.‘ In language, all signs are ultimately arbitrai rior With icons and indexes, both the sign and the object it signifies have— concrete reality. With symbols, the concrete-sign-can_be_used equally well to_ human Tanguage to lie, mislead, exaggerate, and create imaginary worlds in .ovels, poems, or tall tales. We can even speak of impossible worlds as in ience fiction, or of nonexistent ideas, or of phenomena like black holes whose existence we have hypothesized but not proven..None of this would be_ _ possible if the signs of Janguage-were-not-symbolicThis is another important difference between human language and the communicative codes of animals.® Some animals are able to make quite complex indexical messages, but as far as we know, they can only use their signs in the presence of concrete stimuli. Anexample is the dance of the beesan_claborate indexical code by which joes indicate the place where nectar can be found. We know that the dance is__ Fndexical because the bees do not perform it unless they have actually made recent forays for nectar. In other words, they cannot lie about the presence or position of nectar, nor can they invent it. No one knows how the symbolic linguistic code of humans came into being. ‘Some have argued that the origin of language lies in gnomatoposia, that people began taiking by creating iconic signs to imitate the sounds heard - ~ around them in nature. [Link], sometimes called the bow-waw theory, is unlikely to be right, because language in fact makes very little use of iconic. — | words, (Bow-wow itself is conventional. In French, dogs bark with oua-oua, pronounced ““wa-wa.") Another theory is that language was originally, . indexical, arising out of cries of fear, pleasure, and so forth. This theory leaves as much to be desired as the iconic one. Neither theory explains how symbolic signs came to be. We do know that at this point in our biological history our ability to use symbols and learn language is genetically built in, an innate capacity of the human brain. But we know practically nothing about language over ten millennia ago, and very little indeed about language over five millen- nia ago, That is a tiny fraction of the biological history of man, and until we know far more about the neurophysiological aspects oilangusge becan only speculate about the origins of signs as symbols.) System . In general terms, a system is a whole made up of smaller units which stand in particular relation to each other and perform particular Tunctions. ‘For articular relation to eaich other and perform pane.AND LITERARY ANALYSIS 6 LANGUAGE, LINGUISTICS, asa system in which each member is related by particular blood or marriage ties to all the others, and has specific roles and responsibilities. In much the same sense, language 'S ® symbolic system made up of units, functions, and relations. For instance, sounds are units which * combine to make words or parts of words like wn- and -tion, and these in turn are units which can be joined in systematic ways to form larger meaningful sequences, like complex words, phrases, or sentences. In these larger sequences, each of the smaller units has a particular function and particular relations to all the others, Just as the basic sound-meaning combinations are conven- tional, so these larger sequences are highly conventional. For example, in Modern English we say I_saw them, whereas in Old English the equivalent sentence was Ic hie geseah, “I them saw.” employing not only different- sounding words, but also a different word order. At any particular time ina language one will find agreed-upon sound-meaning\félations and agreed-upon orders’-Often there are options—as a system changes from the I iar to the I saw them order, for instance, both structures will exist side by side for some time, with different frequencies over time. They will probably also develop different functions in a speech community, being used for different styles or for different topics. Thus in the sixteenth century, the older word order was largely restricted to poetry. In short, even where there are options in allinguisfic system, they are not random, butare restricted according tocontext. Personal pronouns provide another example of the systematic character of language. In any language, the personal pronouns form a set of interrelated members, each with a different function—a little subsystem, in short. Thi: system is always internally organized, but the way it is organized varies a | . from language to language. In English, for example, we have three third person pronouns (he/she/it) whose function is to distinguish sex. In th i person, English used to distinguish a single addressee (thou) from mo ‘henene (ye), but this distinction is no longer part of the system. Ind sian makes some rather different distinctions. For instance, we in Tndonesian 's anes guished according to whether it excludes or includes th san in distinction we can make in English only by using rathe tresses This & 8 like John and I, we're going to the movies tonight (exclusive of the adidrcosce) Se ae eee ne te menis Caneht (celusive of the addressee), adresse) In the Indonesian syste ne to fie movies tonight {inclusive of the distinctions are made according to degree OF Canale Mede foe sex, but there is both a familiar and an unfamiliar (or polite) poncaa in thee Thus the Modern English and Indonesian pronoun vetene loo epee when set side by side (F stands for Tan nous systems look very different clusive,” In for “inclusive” in th har.” P for “ polite,” Ex for “ex- ver, the diferences between the sets ee ee? of the next e /een the sets are not rand Page). How- system, that is, asa set of units held togethen bo a rote ections as @ The set of organizing principles t tek er by a network of relations. set of “rules.” In linguistics, we speak hae ontrol any system can be called a language as being “rule-governed,”” and we say, for instance, that English has a rule specifying 1 saw him as the Ys > Engl S specifying as instance, a family can be viewedWHAT IS LANGUAGE? 7 Singular Plural Singular Plural Be | i oe ia (F) beliau (P) Modern English Indonesian® normal word order for sentences, and a rule specifying that sex is distinguished in third person pronouns. But this term “rule” may be misleading. In linguis- tics it does not mean that language is controlled by notions of what ought to be said. Rather, it means that everything that is said is organized systematically according to a set of principles internalized in the brain of the speaker. In this sense, “ rule” simply means a pattern observed or observable by the investigator — ~of-language. Another common way of making this pointis to saythatthe rules of language are “*cOpstitutive"* rather-than.. prescriptive.” The rules of etiquette are prescriptive; they prescribe how one should behave. The rules of language, on the other hand, constitute language, the way the rules of poker constitute poker. The linguist’s rules describe how a language is; they do not Prescribe how it should be spoken. Learning to talk involves internalizing the system of a particular language at a particular time, something every child normally does spontaneously. Each of the next eight chapters in this book deals with a different aspect of the linguistic system of English. Chapter 2 deals with the sound system, Chapter 3 with the system of: ‘sound-meaning relations at the level of the word, Chapters 4 and 5 with sentence structure and sentence meaning, Chapters 6 and 7 with the structure of discourse, Chapter 8 with how the English system varies from dialect to dialect. Chapter 9 examines what happens when the English system comes in contact with other linguistic ‘systems. ‘ Language Universals As we have seen, much of language, incluaing the basic sound-meaning relations, is arbitrary and conventional. But not everything is. Linguists have been puzzled for a long time by the fact that the languages of the world are actually much more similar to each other than one might expect. For example, though some languages like Modern English, Yoruba, and Thai favor the word order I saw them, and other languages like Old English and JapaneseLYSIS. a LANGUAGE, LINGUISTICS, AND LITERARY ANA\ favor I them saw, and yet others like Welsh and Hebre fe vor “a hater, ther three logical possibilities: Saw them 1, yey fee ne eet tie does not mean’ that the last three orders are Foon Love, of 1 i i i ances such as = ryt: hy a oe he cl ree phasis, repetition, and so forth. But rarely are the} Cem then, that there is some kind of general restriction or onistrai ‘ * can be a basic order in a language, Pronoun systems exhibit similar restric- tions. For instance, in the languages of the world, we do not nds site that distinguishes familiar versus polite in the first and third person forms but not in the second person forms. Nor do we find a system which uses the same term for I singular and you plural. ; + '-" Facts ike these indicate that not all of language is purely conventional or culturally imposed. Languages are partially shaped by uniyersal constraints on what combinations of elements occur, what historical,eRanges occur, even on what meanings are distinguished in languages.” Some linguists argue that the universal constraii ¢ genetic, part of the human capacity “For language with which one & bone Obes ca Ee general cognitive structures such as our tendency to perceive dualities more Teadily than three-part structures. Still others argue that the universal constraints are explicable in terms of the functions Janguage serves, such as self-expression and manipulation of others. (What would a language be good _ for iTitmade no distinction between I and you?) The issues once more concern the origin of language and are esser;ii: |» unanswerable at the moment, What is important is to recognize that by 2 meaus everything in a particular lan- guage is arbitrary and unique. The universal tendencies in the structures of languages are so widely attested to that at the present time it seems justifiable to claim that part of our cognitive capacity is specifically linguistic. Creativity Recent experiments with chi only learn individual symbol: butcan also learn to combin me key." However, all human speaker. guage without spe: number of element messages, When linguists speak of the “creativity” usually referring to these two ch is able, without g cific ii tie Pe instru mpanzees suggest that certain animals can not is (in this case manual rather than vocal signs),_— ie them in ways reminiscent of sentences like Give asfaras we know, they cannot do certain things which S$ can—they do not appear to be able to | i > “ ¥ ‘ a * cific instruction, and they cannot, on the b na of faa ° asis of a small ts and relations between them, create “ an infinite number of © of human language, they are ‘aracteristics. Anyone who knows a language ion, to prodWHAT IS LANGUAGE? 9 utterances which function in the society as gestures of group solidarity. somewhat like the mutual grooming of monkeys, But obviously, humans are not limited to such routines. The number of sentences possible in a human language is infinite in principle, for there is no limit on how long a sentence can be. It makes no sense to say, for example, that the longest sentence in English (or any other language) is a thousand words long, since, for any longest sentence, someone can propose another even longer. The easiest way to lengthen a sentence is to add parts introduced by and. More complex, but equally infinite structures can be created using subordinating relations, as in I expect you to force Bill to leave. Structures of this sort can be recycled, asin this line from Thom Gunn's poem “Carnal Knowledge,” which theoretically could go on forever: You know I know you know I know you know.® of ‘ To sum up,-the creativity of language consists in this fact: The number of elements and rules in the system is finite, while the number and length tterances the s: is-infiniie. In this respect, linguistic systems are somewhat like the number system. Given any number, one can always construct a larger number by addition or multiplication. In practice,we are limited, of course, by space, time, memory, interest, and many different factors, so that no actual sentence will ever be infinitely long. But what is important is that the system has this potential, Ambiguity Creativity is a linguistic universal, that is, a characteristic shared by all human languages. Ambiguity is another, and is of particular interest to linguists. Ambiguity in language results from the fact that there is not always a one-to-one correspondence between expressions and meanings. For instance, the single sound sequence eal isa sign for a color quality, a kind of stick, and (with the same sound though different spelling) a bucket. Sentences too can obviously be ambiguous, like I_speak to you as a mother. Language differs in this respect from mathematics, which is carefully constructed so that each symbol or sequence of symbols has only one meaning. Teachers and writers often frown on ambiguity, seeing it as a hindrance to communication and 2 symptom of unclear thinking, as indeed it sometimes is. Poets and literary critics often deal with ambiguity as a creative device that concentrates meaning in few words, and it can be this too. Linguists are not interested in evaluating ambiguity, but they are interested in its presence, in the kinds of ambiguity language permits, and in the knowledge speakers have about ambiguity. For instance, taken out of context, the following sentences in English are ambiguous and can thus be interpreted in two or more ways: ‘You may go, All of the members weren’t present, Eleanor took Bill’s coat off, Sam almost killed Pete. Most importantly, linguists are interested in the fact that speakers know sentences such as these are ambiguous and are usually10 LANGUAGE, LINGUISTICS, AND LITERARY ANALYSIS rases that “ disambiguate” the te, one might distinguish the follows: Sam nearly killed Pete bat decided not to and San nearly killed Pete, his driving was so reckless, Secondly, linguists are interested ’ i ttered in context, language- jn the fact that when ambiguous sentences are uttered | : ers are nearly always able to tell which meaning is intended. Accounting for abilities like these is a major task in linguistic theory, and we will return to | this point throughout the book. an important distinction needs to be made between ambiguity and_ vagueness, Ambiguity involves. : scieanings {or one. word, \ phrase, or sentence. Vagueness ithlack of specificity know ‘Some Frenctris- vague because it specifies nothing at all about how much and what sort of French the person knows. But it is not ambiguous. I neither confirm nor deny the report and It’s sort of exciting, but not really are also | vague and lacking in specificity, but they are not ambiguous. There is no more than one distinct meaning for each of them, however vague! that meaning may be, ‘able to explain the am biguity through paraphi meanings. For example, for Sam almost killed Pe [two meanings as © & LANGUAGE AS ACTION IN CONTEXT” Knowing a language means a great deal more than simply knowing how to produce sentences; it also means knowing how to use them. When people speak, they intend things. They speak in order to accomplish something in | what has recently come to be called the “linguistic marketplace,” the inter- actional situations in which language is used to explain, describe, criticize, amuse, deceive, make commitments, deliver opinions, get others to do things, and so forth. From one point of view, sentences like Answer the question or Let’s go to the movies are instruments for producing action in others. From another point of view, the sentences themselves are actions performed by the speaker. To explain something is to do something verbally; to ask somebody a question is to request that they answers. nce: directed actions or acts"tas many different classes of speech acts include utterai fh i people to do things; contractual speech ac's; such as nomen tion agresing on a plan; and describing, informing, and expla sine, wih ee with criticizing, judging, and evaluating, play an importan weir eseene » » and ting, play an important role in education. The speech act of greeting illustrates yet another i vii language serves, that of expressing and establi Ning dee fonction which people. A less obvious function of langua, sear Sarr mriripeine a establish shared worlds between shears mat been is dhe way itis used to friend are arranging furniture and the friend. facing yo a litte t the left. You are probably intended to meve it te he gene which is your right. In other words vou an pmove it to the speaker's left, Share the speaker's perspective. Expressions a eing asked to enter into and Spome, go, bring, take, left, right usually require ghee ne here there, or like | quire the addressee to share or at 1¢ thought of asLANGUAGE AS ACTION IN CONTI least interpret the world in the same why the speaker does, and thus they estab- Nisha shared world of discourse, Ina later chapter, we will be examining more closely the importance of this language function in communicatior Another basic function of language is that of giving names to concepts and things. Nami often seen as a way of getting control over things, For example, thé small child ant naming of objects is usually understood as an effort to establish a kind of naming control over them, This same impulse is reflected in many creation stories, like the first chapter of Genesis where Adam names the beasts of the earth and so becomes their master. Another way of gaining linguistic control over an object is to describe and define it, language used for inti language.” Terms like “experiment,” “entropy,” “chi-square are part of the imetalanguage of mathematics, psychology, and other disciplines, * Sentence,” “sound,” “rule,”. and “ambiguity” are all terms belonging to the meta- ; language of linguistics. Language is unique in that it is not only the medium for: describing everything else in the universe, but also for describing itself. The fact that language can be a metalanguage for itself sometimes causes difficulties, for we often fail to sort out the metalinguistic meaning of a__ term from its other meanings. In the case of “subject,” there is no real problem since “subject” as a linguistic term is now so far differentiated in meaning. from its usual meaning of matter or topic that most people think of it as one word with two meanings. But in the case ofa term like “cause,” we have more difficulties. In the metalanguage of linguistics, all of the following sentences are “causative”: I had her go, I made her go, I got her to go, Bill lightened his load. One could certainly argue whether these sentences really involve causa- tion in its usual sense. A philosopher might claim that such usage is too loose; a layman might find it too abstract or implausible. But in calling these sentences causatives, the linguist is using the word metalinguistically, separate from other usage, to refer to the fact that all the sentences involve an agent_manipulating something or somebody. for_a..certain. purpose. In reading this book, it is important to be able to recognize when language is being used metalinguistically, because we will obviously be relying heavily on the metalinguistic function of language, and will be introducing many terms probably familiar to you, but with different meanings. When language is used, it is always used ina context. What gets said and how if get said is always in part determined by a variety of contextual Tactors. One such factor is “channel,” the medium of expressions being used. “Face- to-face speech, telephone speech, and writing are all different channels, and each constrains choice of expression in particular ways. For example, English has a different range of greeting expressions for each of these channels. On a larger scale, a lecture that is truly appropriate to a live audience that can respond immediately is not appropriate to print, or vice versa, for print is visual and allows the audience to go back or ahead and impose their own temporal scheme on that given by the writer. Without gesture and tone of12 LANGUAGE, LINGUISTICS, AND LITERARY ANALYSIS rd are governed by another contextual factor, ih situation. In speech, I’m going to ke off represent different degrees of formality, and each fh contexts, Similarly, in writing, ediately contrasts with less formal spoken and the written wo namely the degree of formality of the speec' eave and I’m gonna tal corresponds to a different range of sper | | | formal You are requested to reply imm t Please answer right away.'° Another set of contextual factors that intersects with those already. | mentioned has to do with the identity of the participants in communication. “What gets said and how it gets said is partly governed by such factors as the Gea sex of speaker and hearer, For instance, What @ darling pair of shoes js more likely to be spoken by a woman than a man. Men have been shown to interrupt and use direct command forms of the type Do that or Do that, please to a greater extent when addressing [Link] when addressing women.’* | ‘Age differences can be just as important. Speakers of different generations often tend to be more formal with each other than speakers of the same | generation. | The regional, social, ethnic, and educational backgrounds of speaker and hearer are another set of contextual factors that can come into play. A person from a rural town in Kansas may choose to use thi Kansas dialect only when in Kansas, or only to those he or she considers inferior in Kansas, or only with jon7Kansans to aifirm regional identity in the face of the social demand for a “standard” pronunciation. Use of distinct regional forms, as of socially or ethnically distinct markers like ain't, bad (meaning “outstanding, powerful}, or schlepp (meaning ““drag”), fo persons not in the same group often estab- lishes social distance from the addressee and strong affiliation with the regional, social, or ethnic group implied. From the linguist’s point of view, any language is best viewed as a conglomeration of language varieties which differ with respect to the regional, social, ethnic, sexual, generational, and educational groupings within the society. “Standard” languages—like Mi western in the U.S. or Received Pronunciation in Engl: i a veers as one variety among many. meland—are simply seen or a a to some degree determined by a complex set ntextual factors in g who is talking, who is being addressed, essed, what is being taiked about, and others, If s i F about, . If speakers did not i account, all kinds of communicative aisties-would ree a Tenors it ample, a person exclaiming Look at the mess in here! over the tel lawyer addressing a jury with Look, you guys gotta give my « ene bib ent ak, OF a. citizen claimingZo be or not to be, thai , that garbage collection: the question at a town meeting on = Brags colerion. sult. Imagine, for ex- THE TASK OF THE LINGUIST Linguists study the ways in which the so und-meani guages are structured and how they function’ Ac nan ne Corelations of tan- As with all other fields of study,THE TASK OF THE LINGUIST 13 there are a variety of ways to do this. Suppose a linguist were asked to analyze ‘sou eaning Correlation You may go. One possibility would be to take a historical view and to show that while this sentence is ambiguous in Modern English, the equivalent sentence in carly Old English some twelve hundred years ago was unambiguous and meant “You have the strength/physical Power to go” (may is related to might, “strength"), You might show that a weaker meaning of “have the capacity to” developed from the sentence, and that this meaning in turn allowed for the development of the “maybe” meaning, and eventually, some six hundred years ago, of the “permission” meaning. {A historical study would, of course, be concerned with the ways in which these new meanings could arise and the reasons for the changes. Another, rather different possibility would be to consider You may go from tionpunder what circumstances would a contem- porary speal ay go over You have my permission to go, You can go Row, or even some nonlinguistic act like getting up and shuffling Papers for one meaning of this sentence? Under what circumstances would Maybe you'll be going be selected for the other meaning? This is a task involving the — €iicoding'of messages and 1s a task to which computer-simulation of lan guage production is addressed. Another approach is to consider the sentence from the point of view of the hearer: By what means does a hearer know how to disambiguate the sentence in a given context? This is a task (decoding, Yet another approach, the one we will take in this book, is to consider the sentence from the point of view of the code that makes either production or perception possibletn the metalanguage of linguisti ip code that the linguist produces is called a “ grammar.” ? As might be expected. a variety of ways of constructing grammars hay : Bore Impey andere ee hae asa pense — what is found in specific utterance samples, for example, a body of elicited texts Provided by informants."* This is an easy sounding task ; but in practice it is actually so difficult that no one has been able to achieve it satisfactorily, at least where sentence structure as opposed to sound structure is concerned. This is because sufficient examples of any particular structure simply do not occur in any piece of running text. Other linguists, and we share their view, consider that concrete utterance samples are not enough in themselves without access to the intuitions of users of the language in question. These linguists argue that the role of utterance samples is simply to illustrate and test intuitions. For example, any native Speaker of English knows by intuition that there is something odd or wrong about the expression her obviousness to dance, and nothing at all odd about her eagerness to dance. On the basis of this intuition, the linguist extrapolates that obviousness and eagerness do not occur in the same Tange of positions, and goes on to explore this difference further, A corpus of real data could never have revealed this difference, however, because it would contain only actual English utterances, and not impossible or ill-formed ones like her obviousness to dance.u LANGUAGE, LINGUISTICS, AND LITERARY ANALYSIS In the view that we are adopting, intuitions about what is and is not possible in the language are the data of main concern to the linguist, Sample utterances are viewed simply as illustrating the knowledge and intuitions of speakers. Linguists who share this view consider it their ulumate task not only to describe what is found, but also to explain it in terms of internalized | knowledge and innate (genetic) capacities for language, as was suggested | earlier in connection with language universals, These capacities are inferred not only from contemporary languages but also from what we know about | how and why languages change. Part of the ultimate task is to distinguish that | which is universal in language from that which is unique to a particular | language or language-group.'* ‘The question about universal features of language inevitably draws the researcher further into the question of how itis that language-learning takes place in the way that it does, partially indepen- dently of input, and in steps that resemble each other across languages to an extent far too great to be merely random.'® \ Whatever approach they adopt, linguists are in fairly general agreement about the basic subdivisions of linguistic investigation. There is “phonology,” the study of sound systems (from Greek phono-, sound”), “syntax,” the study of phrase and sentence structure (from Greek syn, “together,” and tassein, “ arrange"), and “semantics,” the study of meaning (from Greek semainein, “to signify"). Many_also recognize the area of “morphology,” * basically the study of word formation (from Greek morphe, “form™). Of | more recent date 1s “pragmatics,” the study of language use (from Greek pragma, “deed, aflair"J, Often each of these subdivisions is seen as constitut- ing a separate component of the grammar of a language Many linguists are interested not only in describing languages, but also, in constructing theories, that 1s, well-organized hypotheses about how lan- guage works. Like all theories, whether of physics, biology, or literature, linguistic theories can never be proved right, but must be judged by how much data they can account for, that is, by how many questions they can answer and to what extent ron can show relationships that hold the systems under investigation together Onc theory that has proved particularly success- ful in addressing the problems of what is or is not universal and what makes language a specifically human phenomenon is the theory knownas | E | F a ma “generative, grammar” developed initially by Noam Chomsky in 1957 j TIS“BOOK | ‘Syatactic Structures, and substantially, modified since then Generativ | grammar is a theory of language that aims to characterize what © user knows about his or her language, that is, the knowledge that a | user_puletouse in producing and unders| language | Generative grammar forms the basis for a si foltows in this book; however, we will also be explor hypotheses that have been developed in answer igniticant portion of what ‘ng some of the alternativeTHE TASK OF THE LINGUIST 15 did not pay much attention to, notably the areas of language use and linguistic interaction. In the next few pages, we will outline some of the basic concepts in the approach to language that we are adopting. They are introduced here merely to provide an orientation and frame of reference for thinking about Janguage and about the language of literature. If certain issues are not fully understood from this necessarily condensed outline, this should be no cause for concern since they will all be elaborated on in the course of the book. ‘Competence and Performance. se One of Chomsky's most important theoretical claims ig that we need to distinguish our “* competence,” which Ik what we Know aba ngage our “performance,” which is whal fo when we speak or listen.’7 As lan users, MBWABE (or languages) we speak, we have a kind of linguistic blueprint in our heads, an internalized code, or system. As we have seen in the discussion of You may go, this internalized system is to be distinguished from the activity we engage in as speakers producing messages, or from the activity we engage in as hearers receiving messages. Chomsky sees the linguist’s task as primarily describing competence, and only secondarily performance, because performance is impossible without competence. In other words, the linguist’s grammar describes competence first and foremost. As originally formulated, Chomsky’s idea of competence encompassed onl: e"s abili their language Sut of context, and to distinguish those that conform to the code of the language from those that do not. In other words, it encompassed the issues discussed earlier in the section entitled “What Is Language?" Others have argued, however, that competence also encompasses the ability to use language to accomplish particular communicative goals in particular social and linguistic contexts.’ In other words, they argue that competence also encompasses the issues we discussed in the section entitled “ Language as Action in Context.” This is the view that we are adopting here. An example will illustrate some aspects of competence and help distinguish it from performance. Let us suppose that we are told to analyze the sentence Can you bug the office of the head of HEW? As regards the internal structure of this sentence, any English speaker knows such facts as the relation of the two of phrases to each other, and the significance of the inverted word order can you, which marks the sentence as a question. These observations would therefore be included in our account of the sentence. Our account would also point out that the sen- tence is potentially ambiguous. English speakers know that, depending on the circumstances, this sentence can be a question about the addressee’s physical ability or about the addressee’s administrative power, or about the physical layout of the HEW office. Alternatively, it may be a question about the ability of people in general (you = ‘‘one”). Or yet again it may function as a com- mand to actually do the bugging. We will also have to point out that if it is: NALYSIS 16 LANGUAGE, LINGUISTICS. AND LITERARY Al “well-formed” (to use some linguistic jargon) 16 command such things. If we were account~ f this sentence, On the other hand, we would have to ask such questions as whether, in this particular case, i ere _ as a question ora command; and if it was 4 command, ia rte cnke och aaa ee he particular speaker actualy did have the authorl) 20 SU Bete command. Clearly, performance takes competence into account, itis not thing. the same thing. ess to competence. We make Just as clearly, we do not have direct acc hypotheses about it from indirect evidence, such as available utterances and intuitions. Some of the pitfalls of this procedure are considered in Chapter 8. For the moment, suffice it to say that a particularly useful procedure for arriving at aspects of linguistic competence is the deliberate manipulation of language to create not only sentences that conform to the rules of the lan- guage but also sentences that do not. For example, we can construct the set He is an entomologist, isn’t he?, They are entomologists, aren’t they?, "He is an entomologist, aren't they?, and She is an entomologist, isn’t he?, and verify that English speakers find the last two peculiar and the first two not. On the basis of such intuitions we can formulate a rule which in essence says that in the grammar of English the pronoun in the isn’t phrase must be the same as the subject of the sentence. Thus, rather than excluding ill-forrhed sentences from the inquiry, the generative grammarian uses them to gain access to compe- tence. In the metalanguage, sentences that are in accord with the system internalized by the language-user are called “grammatical”; sentences that are not in accord with that system are called “ungrammatical” or “deviant.” Ungrammatical sentences are starred, as, for example, *He is an entomologist, doesn’t he? The notion of “grammaticality” that linguists use is descriptive, not prescriptive. A “grammatical sentence” is one that occurs in the system and’is accepted as a norm by the speech community in which it is used. In this sense an utterance like I ain't goin’ is ungrammatical for some English-speaking speech communities (mainly those which use what is known as “standard” English), but grammatical for many others. The descriptive concept of ey tel ioe ae from the prescriptive notion of “what you ought oe a bs Ee bab y came ace in school. Prescriptive grammars SHER eit speak an i in order to become a well-accepted member tend to [Link] least some renee grammars are conservative and Jong ceased to exist i notions of linguistic good behavior that have -xist in everyday language, even of the “ best speakers.” For example, It is I, insisted on in the eighteenth century on th in, is still sometimes insisted on in the school ronithewmodeh ofan! cluding the teacher, actuall chools, although practically no one, in- contemporary change that has ys Teis I, but rather It’s me. An example of # mars is the use of good for as not yee been recognized by prescriptive gram-' prescriptive o idverb well, as in You sang good. While the grammarian would insist on well, the lingui: i , guist would recognize to be used as a command, it is only if the speaker has authority t ing for a particular performance o|THE TASK OF THE LINGUIST 7 that good is often used instead, and would note that well is, dete 7 for many speak restricted to writing or formal speech," Y speakers, - ot ‘ ; ' Underlying anw/Surfate Structure *' tt Fen As we have seen, there are many types of grammar. But every grammar must account for the fact that there is no direct one-to-one correspondence between meaning and sound, The sound a means nothing in itself, but conventionally combined with certain other sounds it can be part of a sign, as in pat, lamp, at. Equally inescapable is the fact that a language is not a dictionary. Signs alone do not makeuplanguage—theyare arranged in sequences, and these sequences involve a hierarchic structure or grouping. Thus we divide up The box on the table is mine into (The box on the table) (is mine) and not into (The box) (on the table is mine) or (The box on the) (table is mine). We divide up the ambiguous sentence I ran into the girls with the flowers into 1 ran into the girls -(with the flowers) and I ran into (the girls with the flowers). This hierarchic structuring, or “putting together,” is somewhat similar to the bracketing of math: 2 x 2 — | is understood by convention to be grouped hierarchically as (2 x 2) — 1; the other meaning of this sequence must be indicated by parentheses—2 x (2 — 1). In a grammar, we need to distinguish at least between sounds, words (and word parts), groups of words, or “ phrases,” like on the table, and sentences. Describing sound-meaning correlations is not always a straightforward task of dividing sentences into various kinds of groupings. Ambiguou: sentences present one kind of problem. Consider, for example. The fish is ready to eat, which is ambiguous as to whether the fish is going to eat or to eaten. Do we say in this case that there is one sentence with two meanings o| two separate sentences 7. How does the grammar correlate this sound sequence with each of its two meanings? Generative grammar answers this question by positing two levels of sentence structure: surface structure” and a more abstract structure which we will call ‘underlying structure.” The underlying structure unambiguously specifies the basic meaning and categories of the * sentence: this structure is modified in various ways to become a surface structure, which is the linear arrangement of _words_and_phrases which will be pronounced. The fish is ready to eat is a surface structu In one of its interpretations, this surface Structure corresponds to an underlying struc- ture something like The fish is ready (for someone (o eat the fish). In the other interpretation, it corresponds to the underlying structure ‘The fish is ready (for the fish to These two underlying structures can. fie for these modifies s “ transfarmations. ; ; Sars rmarion between surface and abstragt or underlying structures is one of the most basic and innovative concepts of generative grammar. While this distinction is accepted in ‘most current linguistic thinking, there is118 LANGUAGE, LINGUISTICS, AND LITERARY ANALYSIS disagreement on what the underlying structure should look i a hows asta © it should be, and what information about the sentence it oe include, ty, the course of this book, we will be looking at a number of if rent views of underlying structure. Eor the moment, let us repeat only that it ine tudes both the basic meanings and the basic syntactic categories 0! the sentente, 80 thar Sentence looks something like an extended para ~ hrase of the surface sentence. Surface structur: - be derived from vindertying structures by the application arapstormational ules. As we shall ee ea Reeve gimmar aso includes a set of “ phonological | rules” which modify surface structures into sentences as they are actually pronounced. A Model of Grammar In linguistic metalanguage, a grammar is seen as consisting of various components. The underlying syntactic/semantic structure forms one compo- nent, the surface syntactic structure forms another, and the surface phonetic structure yet another. Sets of rules state the relation of one component to another. For example, rules relate the underlying structure The fish is ready (for someone to eat the fish) to the surface structure The fish is ready fo eat, and rules relate the latter to the actual sounds (consonants, vowels, rhythm, rise and fall of the voice) that represent this sentence in speech. As we mentioned earlier, in the approach we are adopting, and in accord with much current linguistic thinking, a model o1 competence must also account for the language-user's ability to use language in relation to specific contexts and specific communicative goals. These are described in what is called the “ pragmatic” component of the grammar. The relations between various components of the grammar can be modeled as follows: Underlying syntactic) semantic structure Pragmatics 1 Transformational rules |_ Sees oma re Phonological rulesAPPLICATIONS TO LITERATURE, 19 : The arrows indicate dependency relations—some phonological facts, ‘espe-—. cially pitch patterns, for example, depend on syntactic structure. In turn, choice of sentence structure and pronunciation as well as choice of what is said at all depend on contextual factors specified in the pragmatic component. In subsequent chapters, each of the four components will be discussed in considerable detail, starting with the phonological one because it is in some. ways the simplest to describe and because it can be used to introduce nearly all the basic linguistic concepts. To set the scene for our dual purpose of dis- cussing, both what the study of the English language involves and also what it can contribute to analysis of English literature, first we must describe in a general way how the broad linguistic hypotheses introduced in this chapter can be relevant to literature. ' “APPLICATIONS TO LITERATURE “Applying” Linguistics When people talk about “applied” fields, like applied linguistics or applied physics or applied anthropology, they usually refer to-the use of the discover- ies, the frame of reference, and the terminology of one discipline to serve the ends of another area of endeavor, The [Link] methods of the social . scicnoes are often applied to the solution of Santee sci problems. Many applications of linguistics are pra way. The study of speech sounds and the physiology of speech, for example, have been of use in therapy for the deaf or those with impaired hearing. Semantics and discourse analysis have contributed to the understandirig and treatment of aphasia (loss of speech through brain damage). Contemporary phonology and syntax have markedly influenced methods of language teaching, both of first and second languages, while the study of discourse and linguistic interaction has provided new insights for workers in psychotherapy. Research in dialects, bilingualism, and the interaction of language and class structures is applied by governments in formulating their national language policies and conducting the language planning so important to the many multilingual nations of the world. At the same time, computer scientists rely heavily on linguistic insights in their work ‘on machine translation and on computers which can produce and recognize speech, Pragmatics and the study of speech acts have recently come to the fore in the area of language and the law, where they are applied to such questions as truth in advertising, definition of libel, and interpretation of the law itself. The study of style and the language of literature is one of the most traditional applications of linguistics, one which has been given new impetus by the rapid new developments in linguistics since the development of genera- tive grammar. At the present time, linguistic analysis of literature is one of the20 LANGUAGE, LINGUISTICS, AND LITERARY ANALYSIS most active and creative areas of literary studies. As is the case with its other areas of application, linguistics is not essential to the study of literature, Certainly one does not need to know linguistics in order to read and under. stand literary works; and criticgl analysis has long been carried out without formal linguistic apparatus,” However, linguistics can contribute a great | deal to our understanding of a text. It can help us become aware of why it is that we experience what we do when we read a literary work, and it can help us talk about it, by providing us with a vocabulary and a methodology through which we can show how our experience of a work is in part derived from its verbal structure, Linguistics may also help us solve problems of interpretation by showing us in tigorous ways why one structure is Possible but not another. Above all, however, linguistics can give usa point of view, 4 way of looking at a text that will help us develop a consistent analysis, and Prompt us to ask questions about the language of the text that we might _ otherwise ignore. Since texts are the primary data for all literary criticism, adequate means of textual description are essential if any criticism is to be Properly founded. Linguistics helps ensure a proper foundation for analysis, by enabling the critic to recognize the systematic regularities in the language ofa text. In fact, we can u: Serovcenstensta theory about the language of a textin the Torm_of al‘ grammar of the text."lin this sense, although itis relevant to all criticism y Hinguistics does not encompass {i ferary criticism, Literature as a Type of Discourse Though we sometimes tend to think of literature as a realm of free, individual expression, it is in many respects highly conventionalized, like everything else in language. One important set of conventions are those governing literary genre. In linguistics, the term “genre” is used to refer not only to types of literary works, but also to any identifiable type of discourse, whether literary or not. In this sense, the lecture, the casual conversation, and the interview are all genres, just as the novel and short story ace. This broader. view of genre is valuable in that it helps us conceptually to bridge the traditional gap between literary and nonliterary discourse. It enables Us to view literature as a particular range of genres or discourse types, that is, as a particular subset of the repertory of genres existing in a given Speech community. In our own culture, there is some disagreement about exactly which genres constitute literature. There is little consensus, for example, on the status of the limerick or the nursery rhyme; the distinction sometimes drawn between literature and folklore is dubious at best; a religious poem might be considered literature when it appears in a poetry anthology, but not when it appears in a hymn book. For the purposes of this book, we will be using “‘fiterature” in the sense that it usually has in the phrase “modern English literature,” that is, novels, poems, short stories, and so forth, Among characteristics of literature as a Tange of genres is that it is gener-APPLICATIONS TO LITERATURE, ui ally public, not private, discourse, In addition, written literature is discourse + that may be read at a far distance in time and place from its origin. This means that the relationship that holds between speaker/author and hearer/reader in written literature is of a very special sort and one that is a particularly im- portant aspect of what can be called “literary pragmatics.”?? Furthermore, literary discourse is often fictional. One of the pragmatic conventions of fictional narrative is that the speaking / of the speech act is understood not to be the author of the work, but an intermediate narrator or addresser who has been created by the author. Within the fictional world of the story, the narrator (or addresser), not the author (or speaker), is held immediately responsible for whatis said. Thus we speak of reliable and unreliable narrators, not authors. The author is responsible as speaker, however, when the fiction comes to be judged by the external world, as in criticism, libel suits, censorship cases, or Pulitzer Prizes. Aside from genre conventions, literary discourse has many other general linguistic characteristics for which the linguist can provide tools of analysis. Certain kinds of phonological, syntactic, and semantic phenomena occur with much greater frequency in literature than in other kinds of discourse. For example, “poetic” devices like metaphor, alliteration, and archaism are commonly associated with literature, although they are, of course, not unique to it. The conventions of rhyme and meter constitute elaborate formal con- straints on phonology, syntax, and vocabulary, and the study of grammars can help show exactly what these constraints are. One of the most important characteristics of literary discourse is its recurrent linguistic patterning, or “cohesion,” a patterning which may be found to operate at all levels of the grammar; and it is here especially that the linguist can throw light on the language of a text, demonstrating both what the linguistic system in the work is and how it operates in that particular text. Cohesion: ~ a . - smut - = a tee = RE HOM + The idea of cohesion was first developed in detail by Roman Jakobson, one of the leading linguists of the twentieth century and a pioneer in the applica- tion of linguistics to literature. In 1960 Jakobson characterized, with reference to poetry, a notion basic to the analysis of literary texts: that they have Cohesion or internal patterning and repetition far exceeding that of most non- literary texts.2° Cohesion in poetry is usually discussed in terms of repeated refrains, regular stanzas, rhymes, alliteration, meter, and similar devices, Jakobson’s interest lay not so much in these well-known features but in rather less frequently discussed linguistic features, especially linguistic cohesiveness created between elements at different levels of the grammar, such as parallels * between meaning and sentence structure, or between sentence structure and Sound structure (and, of course, their interplay with other specifically Postic \ features, such as meter),2 LANGUAGE, LINGUISTICS. AND LITERARY ANALYSIS jakobson describes the phenomenon of cohesion as feito “The tic function projects the principle of equivalence from the axis of selection in he axis of com wav A difficult sentence, but probably the one into the axi ature. What he means by cK i roaches to liter: most often quoted in li Q ie, What this is that, in poetry, structures which are roughly equivalent in sound, or sentence structure, OF grammatical category, or some other aspect tend to be combined in a linear ‘order or sequence. Poetry, In other words, involves partial repetition, whether of metrical patterns, rhymes, or sentence structures, Jakobson cites Caesar's famous veni vidi, viel as an example. This sentence combines in sequence three words of the same grammatical category (verbs), same inflection (first person singular past tense), same number of syllabies, nd structure (rhyme and alliteration). same stress pattern, and very similar soul This extraordinary cohesiveness is what makes the sentence so memorable. In the English I came, I saw, I conquered, some of the effect is lost because of the s versus k, and the two syllables of conquered versus the single syllables of the other words, but the sentence is still strikingly cohesive. At the semantic level. the cohesion has a particularly interesting effect. By seeming to equate the acts of coming, seeing, and conquering, Caesar's sentence implies that the last act was as easy for himas the first two. Hence, the im) i jestic arroy it produces. Political slogans and ‘advertisements thrive on the principle of cohesion,— in part because it makes them easier to remember. For example, among advertisements we find Tarn on Schick, turn out chic for a Schick hair styler and Silk and Silver turns gray to great for Silk and Silver hair coloring. These both involve cohesion in phonology, vocabulary, and meaning associations. In poetry the principle is usually exploited in a more subtle way; but cohesion is present at least to some degree, except in some experimental poetry that deliberately rejects it. ‘A relatively simple literary example of cohesion is provided by Robert Browning's well-known song from “Pippa Passes”: nbinatior inguistic IPP! The year's at the spring And day's at the morn; Morning's at seven; The hillside’s dew-pearled; The lark’s on the wing; The snail's on the thorn: God's in his heaven— All's right with the world 13° Patterning is evident at every level in this poem, and at the same time we ca ee phat patterns are varied to avoid monotony. One of the most obvious ive patterns here is the syntax—cach line is a single clause consisting ofAPPLICATIONS TO LITERATURE a noun + 's + X, and X consists of a prepositional phrase everywhere except in lines 4 and 8 where the pattern is varied. (But note that this variation itso is cohesive, in that it occurs at the same point in each pair of four lines, and involves an adjective in both cases). The syntactic cohesion coincides with semantic cohesion among the nouns. In the first four lines, we find 2 series of time nouns in order of increasing specificity: year, spring, day, morm, morning, seven, joined by the preposition at. This extremely tight patterning is loosened and played upon in the second four where lark and saail ace linked both by semantic likeness (two animals) and contrast (“ higher” verses “‘lower™ animals). With on the wing and on the thorn, syntactic and lexical likeness interplay with semantic difference (on has the same form but dif- ferent meaning in each case). In line 7, God is forcibly incorporated imto the pattern and thus unexpectedly placed on the same levei of existence as lark and snail, while the change of prepositions from on to in keeps the parallelism from being complete. Metrical patterning in the poem interacts rather subtly with lexical and syntactic patterns. In each part, the first, second, and fourth lines have the pattern ~/-~/ (where / means a stressed syllable and - an unstressed ome), while the third line has the mirror image /~~/-. Thus, though the last lime of each part breaks syntactic and lexical patterns, it does conform metrically. while metrical variation is used in the third and seventh lines and heips counteract any monotony arising from the syntactic and lexical cohesiveness. At yet another level, the internal regularity of this poem is itself coumtex- balanced by the fact that, with respect to normal spoken English. many of the poem's expressions are decidedly “irregular.” We do not say that years are “‘at the spring” or days “at the morn.” These expressions illustrate the fact that literature often uses expressions that are not, or are no longer, common in spoken conversational language. Moreover, in this poem, the fictiomal singer, Pippa, is a young Italian woman, and the song, fictionally, is being samg in Italian, Doubtless, Browning wants to remind us of this by estrangimg us from the English in the text. The phenomenon of cohesion in literature obviously has everything to do with the fact that literature is art, that literary texts are constructed to produce in us the kinds of experience we speak of as “aesthetic,” im whack symmetry and interplay of sameness and difference play a major role A complete understanding of cohesion will depend on further understanding of aesthetic experience and perception. . f the secrets of good poetry is to be cohesive, and yet not too muck One of so. Tao Tigi an equivalence leads readily to doggerel. While the literary cxitic may want to evaluate the cohesiveness in a text, the linguist’s task is mot so much to evaluate as to demonstrate exactly what is present and what is not. It is also the linguist's task to show where there are differences as well as similarities, or where there is some variation in a pattern or some ind of opposition between surface and underlying patterns. We will be examiningANALYSIS 4 LANGUAGE, LINGUISTICS, AND LITERARY inguistic poi fiew in this a veral mons examples of cohesion from a linguistic point of vi nd se eo the next two chapters. ~The Idea of the “Grammar” ofa Text __ i i i tematic From the point of view of generative rammer, discovering the aie em ate regularities in the language of a text is only a partial step tofatextn of the text’s linguistic structure. Rather than limiting an accoun ; ‘ observed regularities, we can go on to make generalizations about tl etext s phonological, systacti ind semantic structure and its Pragmatic charac. teristics, gencralizations which will reveal the text's stylistic traits and tenden- cies and the principles on which it is structured. Such generalizations resemble those that constitute the grammar of a language. It has therefore become customary to extend the term “grammar” “grammar of a language” to “ grammar of a text” Const & fof & Ext; then>isa-way of hypothesizing about its overall internal structure. It_enablés the-critic to make stylistic ObservaHiOne in an organized wa way about the most detailed facts inguage. es *—One of the most interesting ways to use a grammar of a text is to compare it with the overall grammar of the language in which it is written. Such com- parison can reveal, for example, what grammatical categories and options have and have not been used, and in what ways the text departs from normal usage. Indecd, a grammar of a text Provides the only rigorous basis for com- Paring the language of an individual text with “the language as a whole.” An example should cla fy the broad outlines of what might be done with the grammar-of-a-text approach. The details can be filled in as you read the following chapters, The text 's T. S. Eliot's “Rhapsody on a Windy Night,” one of the Prufrock Poems, published in 1917, Here the Speaking “1” describes a fong late-night walk and the inner, mental experiences that accompany it. RHAPSODY ON A WINDY NIGHT Twelve o'clock, Along the reaches of the street Field in a lunar synthesis, Whispering lunar incantations Dissolve the floors of memory And ali its clear Telations is divisions and Precisions, very street lamp that I pass Beats like a fatalistic drum, And through the Midnight shakes the me hell ‘$@ madman Shakes a dead 8eranium,APPLICATIONS TO LITERATURE 2s Half-past one, The street-lamp sputtered, The street-lamp muttered, The street-lamp said, “ “Regard that woman Who hesitates toward you in the light of the door Which opens on her like a grin. You see the border of her dress Is torn and stained with sand, And you see the corner of her eye ‘Twists like a crooked pin.” The memory throws up high and dry A crowd of twisted things; A twisted branch upon the beach Eaten smooth, and polished As if the world gave-up The secret of its skeleton, Stiff and white. A broken spring in a factory’ yard, Rust that clings to the form that the strength has left Hard and curled and ready to snap. Half-past two, The street-lamp said, “Remark the cat which flattens itself in the gutter Slips out its tongue And devours a morsel of rancid butter.” So the hand of the child, automatic, Slipped out and pocketed a toy that was running along the quay. T could see nothing behind that child’s eye. T have seen eyes in the street Trying to peer through lighted shutters, And a crab one afternoon in a pool, An old crab with barnacles on his back, Gripped the end of a stick which I held him. Half-past three, ‘The lamp sputtered, The lamp muttered i ‘The lamp hummed: “Regard the moon, La lune ne garde aucune rancune, She winks a feeble eye, She smiles into corners. She smooths the hair of the grass. the dark., LINGUISTICS, AND LITERARY ANALYSIS % LANGUAGE, The mvon has lost her memory. A washed-out smallpox cracks her face, Her hand twists a paper rose, That smells of dust and eau de Cologne, She iy alone 60 With all the old nocturnal smells That cross und cross across her brain.” The reminiscenee comes Of sunless dry geraniums And dust in crevices, Smells of chestnuts in the streets, And female smells in shuttered rooms, And cigarettes in corridors And cocktail smells in bars. The lamp said. J “Four o'clock, Here is the number on the door. Memory! You have the key The little lamp sp Mount. The bed is open; the tooth-brush hangs on the wall, Put your shoes at the door, sleep, prepare for life.” ads a ring on the stair, The last twist of the knife.2* In “Rhapsody on a Windy Night,” the “I” of the poem experiences, as he walks, a mental struggle betwecn two views of the world, one active, focused on the moving present, and one passive, focused on the inert past. This struggle is expressed through two images, that of the street lam) andithat of memory. The poem alternates between their two perapective, tat stanza 2 deals with the lamp, stanza 3 with memorys lines 33-37 wich che lamp, lines 38-39 with memory; lines 46-61 with the lamp, lines so ah is mermory; and stanza 6 with the lamp. The rest of the pocm, stance 1 en 40-45, and the final one-line stanza, consists of direct comm: tt 5b: ned of the poem. (It is surcly no accident that these sections occ! rari ee jag middle, and end.) A detailed look at the language ca ftw that wiht ae overall grammar of the poem there are two sub. romaine than he ee linguistic systems which, although not entirely complementary ae nein The language used in connection with the lamp igre ere Mea 80. used in connection with memory. The fg mt igradically different from that combine in the sections of commentary by the joes ofthe twoiparts What is imy it F iS Portant in the famp's grammar is that action verbs and dit irec-APPLICATIONS TO LITERATURE. ras tional adverbs are so frequent,that everything seems to be in motion, especially individual parts of the body: the corner of her eye twists like a crooked pin (21-22), she smiles into corners (53), her hand twists a Paper rose (57). In the memory S grammar, however, there are static pictures expressed by nouns coupled with Place adverbs: a twisted branch upon the beach (25), = broken spring in a factory yard (30), dust in crevices, smells of chestauts in the streets (64-65). Even those few verbs in the memory part that are usually action verbs have no action meaning here. Thus, in the reminiscence comes of sunless dry geraniums (62-63), memory does not really come in any active sense, but just happens. The two grammars are complementary, each using the vocabulary of the other in terms of its own structure: smell is a verb in the lamp’s grammar (that smells of dust and eau de Cologne, line 58), smell is a noun in the memory’s (female smells in shuttered rooms, line 66); the lamp’s grammar has twist and smooth as verbs (22 and 54), while memory’s has twisted and smooth as adjectives (24 and 26). The fifth stanza provides a highly consistent example of the contrast between the two grammars, the lamp’s expressing the world of action and the present, the memory’s express-, ing the world of stasis and the past. The same contrast is apparent in the pragmatic aspects of the two gram- mars. In the lamp’s grammar, direct speech is possible. The lamp’s descrip- tions of the world come as speech acts addressed directly to the “I” by the lamp. Moreover the speech acts performed by the lamp are predominantly commands in which the lamp directs the “L""’s attention to immediate details. In the grammar of memory, there is no direct speech and no imperative speech-act function. Language is used only for description, not direction, and the perceptions of memory are presented as purposeless, random, and non- verbal in origin. Language is used to describe memory, but language does not exist within the world of memory. Hence there is no metalanguage in memory’s grammar, while the lamp’s grammar is full of it (sputtered, muttered, said). As a verbally active entity, the lamp is able to use the language to establish a shared world with the “I” of the poem, and its grammar includes many expressions with this function. For instance, in Regard that woman who hesi- tates toward you in the light of the door (16-17), that requires the addressee to share the lamp’s spatial domain, in which the woman is distant; in toward you, the lamp adopts the ‘“I'"’s spatial perspective to describe the direction of the woman’s movement. By contrast, there are no pointing expressions establishing an immediately shared spatial world between the “I" and his memories. We do not get, for example, that twisted branch up on the beach over there, Paradoxically, then, the ‘I""'s own memories seem to engage him less than the external and social world pointed out to him by the lamp. . , Despite what one might expect, the two contrasting sub-grammars in “Rhapsody on a Windy Night" do not rule out the possibility of cohesion in the poem as a whole.-The regular alternation between the lamp and“a LANGUAGE, LINGUISTICS, AND LITERARY ANALYSIS memory is, in itself, a source of cohesion, as is the regular pacing by clock time (twelve o'clock, lelf-past one, and so on). In addition, there is a great deal of syntactic cohesion, such as repetition of sentence patterns as in lines 14-16 and 47-49. The basic structure here is article-noun-verb (The street Jamp muttered), with an added adverb, in the dark, at line 48, to prevent monotonous over-repetition. Similarly, there are patterns parallel to She winks a feeble eye (52-53) and to dust in crevices (64-68). More complex examples of syntactic cohesiveness can be found throughout the poem. For instance, there are patterns involving directional adverb... subject-verb- object in the first stanza. Along the reaches of the street... incantations dissolve the floors of memory (2-5) parallels through the spaces of the dark midnight shakes the memory (10-11). Cohesion in the vocabulary is very clear both within stanzas, for example, Held in a lunar synthesis (3) and Whispering lunar incantations (4), and between stanzas (cf. especially street lamp, memory, dark). Semantically, we have a great many words denoting destruction or decay: torn, stained, crooked, broken, rest, rancid. Phono- logically, we cannot miss echoes. For example, in the first stanza there are incantations . . . decisions . . . precisions; reaches... street. Of course, constructing 4 grammar of a simple text is not exactly the same thing as constructing a grammar of a language. Grammars of languages have to account not just for the available data in a language, but also for the poten- tially available data, the infinite number of sentences that could be produced — in the language. Because they have this predictive power, grammars of lan- guages can always be tested against new data in the language. With individual text grammars, there are no other potentially available data against which the analysis could be tested. Still, in theory, a grammar of a text would enable us to characterize a text specifically enough to produce another text stylistically indistinguishable from it. In practice this is seldom completely possible, because literary works are often highly individualized and because their internal structure depends on factors other than purely linguistic ones. But the Possibility that one can know the structure of a text well enough to produce something similar is one on which pastiche and caricature operate, and the idea that a grammar can characterize the language of the text in full detail underlies computerized poetry, since the program which produces such poetry is a grammar of the text it produc: - . Describing a text in terms of its grammar involves viewing it as a single whole in which certain structures are simultaneously present. Likewise, to speak of cohesion as the presence of internal patterns is to treat the text as an object possessing certain properties, rather than as a temporal! progression. Such a view is necessarily incomplete, for a text is also a temporal sequence. A written text is laid out in space because expressed language is a temporal thing. The act of reading takes time; the experience of a literary work is temporal. So of course is the author's act of writing. There are critics whO maintain that the only valid way of approaching a literary text is throug description of the temporal experience of reading.?” Others claim a text mustAPPLICATIONS TO LITERATURE 29 be viewed as a dynamic structure constantly transforming itself as the work progresses.?* In fact we need all these perspectives. [~ Style as Choice | Style as Choice | It is customary not only to talk about the language’ of a text or author, but more specifically about the “style.” People speak of racy, pompous, formal, colloquial, inflammatory, or even nominal and verbal styles. How can such generalizations be characteriza Linguistica! De characterized linguistically? One way is to say that style results from a tendency of a speaker or writer to consistently choose certain structures over others available in the language. With this view, we can dis- tinguish between “‘style” and “language” by saying that language is the sum total of the structures available to the speaker, while style concerns the charac- teristic choices in a given context.2® In this sense, in writing a grammar ofa text, we will probably be more concerned with the style than with the language, for it is not so much that every possible structure available is interesting, as those which dominate a text, or part of it. To claim that style is choice is not, of course, to claim that it is always conscious choice. Indeed, if one had to make all phonofogical, syntactic, semantic, and pragmatic choices consciously, it would take 2 very tong time to _say anything at all. In literature, as in all discourse, a sense of the “best way of putting something" can be intuitive or conscious; the reguit as far as the reader is concerned will be miuch the same. Stylistic choice is usually regarded as a matter of form or expression, shat is, as choice among different ways of expressing an invariant or predetermined content. But this view is misleading, for writers obviously choose content too. In our grammar, with its semantic and pragmatic components, both content and expression can be viewed a5 matters of choice. Choice of content involves choice of semantic structures; choice of. expression involves choice of pragmatic functions and contextual features (such as what relation a speaker adopts toward the hearer, what inferences are to be conveyed, what assump- tions made). Choices in both these components of the grammar are in turn the basis for phonological, syntactic, and lexical choices. This approach provides us with a new way of thinking about whether there is or is not a duality between form and content. This issue has been discussed in philosophy and aesthetics for centuries. A large number of critics and stylisticians acknowledge such a duality, saying that given some particular content (“meaning”) a variety of surface forms are possible. In this view it is possible for there to be sentences that are synonymous, even though they have different forms. The opposing position is that every difference in form brings a difference in meaning and that synonymy is therefore impossible. Now it is clearly useful to say that My twenty-three year old brother is « bachelor is synonymous with (has the same meaning as) My twenty-three year old brother fs unmarried. Yet they are not exactly equivalent, For example, one would scarcely say the first sentence to a0 LANGUAGE, LINGUISTICS, AND LITERARY ANALYSIS hile one might say the second. child because bachelor is a technical term, ¥ e : The difference is pragmatic, not semantic. In terms of our grammar, choosing between these two sentences, the speaker makes a pragmal contextually motivated choice between two semantically equivalent surfape forms. Thus the grammar allows for synonymy (thereby maintaining a form/ content distinction) and at the same time accounts for the fact that synony- mous surface forms are not exactly equivalent. The interplay between semantic sameness and pragmatic difference is exploited by Samuel Beckett in the following passage from his novel, Murphy. In-the climactic scene of this novel, the ill-adjusted and alienated protagonist, who works as an orderly in a mental hospital, looks deep into the eyes of a__ singularly withdrawn patient, Mr. Endon, and encounters nothing but— his own reflection, Endon's complete unawareness of his presence profoundly shakes Murphy's confidence in his own existence. We read: Kneeling at the bedside, the hair starting in thick black ridges__ between his fingers, his lips, nose and forchead almost touching __ Mr.'Endon's, seeing himself stigmatised in those eyes that did not see him, Murphy heard words demanding so strongly to be spoken that he spoke them, right into Mr. Endon’s face, Murphy who did not speak al all in the ordinary way unless spoken to, and not always even then. “the last at last seen of him bimself unseen by him 10 and of himself” A rest. “The last Ms. Murphy saw of Mr. Endon was Mr. Murphy-—| unseen by Mr. Endon, This was also the last Murphy saw of Murphy.” A rest. “The relation between Mr. Murphy and Mr. Endon could not have been better summed up than by the former’s sorrow at__| seeing himeeif in the latter's immunity from seeing anything but himself.” 20 A long rest. “Mr. Murphy is a speck in Mr. Endon’s unseen.” < That was the whole extent of the little afflatulence, He replaced Mr, Endon’s head firmiy on the pillow, rose from his knees, left the cell, and the building, without reluctance and without relief.2? Here Beckett gives us several different formulations of the same rather complic: ted content. The first, highly elliptical, is set out like a poem; the second is an explanatory paraphrase of the first using complete sentences and no pronominalization; the third, reminiscent of nineteenth-century novelistic style, adds some emotive content (the former's eorzow), which is removed in the final aphoristic summary of the situation, where 2 metaphor (a speck in Mr, Endon’s unseen) is introduced. Notice too the way this passage dramatizes