How Language Works
Students studying linguistics and other language sciences for the first time often have misconceptions about what they are about and what they can offer them. They may think that linguists are authorities on what is correct and what is incorrect in a given language. But linguistics is the science of language; it treats language and the ways people use it as phenomena to be studied much as a geologist treats the earth. Linguists want to figure out how language works. They are no more in the business of making value judgments about people's language than geologists are in the business of making value judgments about the behavior of the earth.
But language is a cultural phenomenon and we all have deep-seated, cultural ideas about what it is and how we ought to use it, so knowing where to begin in studying it scientifically is not a trivial matter at all. Issues arise that would not if we were geologists figuring out how to study earthquakes or the structure of the earth's crust. For this reason, before we dive into the study of language, we will need to examine some of the biases that we all have concerning language and to set some ground rules for how we are going to proceed. Because there is more than one way to begin, it will also be useful to establish a basic stance to guide us. Finally, because human language is an enormously complex subject, the book will focus on a narrow range of topics and themes; there will be no pretense of covering the field in anything like a complete fashion. This first chapter is designed to deal with these preliminary issues. But first, you will need to know about the various conventions that I will be using in the book.
The organization of this book is based on the idea that human language has a small set of basic properties, each of which plays a role in the workings of language as an instrument for communication and thought. Each chapter in the book (after this one) introduces a new property. Chapter 2 discusses words and word meaning. Chapter 3 discusses phonological categories, the units that are combined to make word forms. Chapter 4 discusses phonological processes, the ways in which the units of word form interact with one another. Chapter 5 discusses compositionality, the principle that allows complex meanings to be expressed by combinations of words. Chapter 6 discusses how words are organized into larger units and how these allow us to refer to states and events in the world. Chapter 7 discusses how the grammars of languages divide the world into abstract conceptual categories. Chapter 8 discusses the productivity and flexibility of language and how grammar makes this possible.
Most linguistics texts draw their examples from an unconstrained set of languages. This has the disadvantage that students are left with little sense of how the different aspects of each language fit together. It also invites the kind of errors that may crop up when linguists rely on examples from a wide variety of other linguists. For these reasons, almost all of the examples in this book are limited to a set of nine languages. You can see the word for 'language' in each of these nine languages in the upper-left corner of the Table of Contents page, and, together with words in eight other languages, at the top of each page. If you're interested in knowing more about these languages, each is described briefly in this appendix.
Throughout the book I will include links to other references that are available online. In particular I will often link to articles within the English edition of the collaborative encyclopedia, Wikipedia. Wikipedia includes some articles that have not been written by knowledgeable people or have suffered from disagreements among the editors, but for important topics on language, the articles have been edited many times and have stabilized into relatively useful and reliable overviews of these topics. Because I will link to Wikipedia so often, I will use this special symbol with the Wikipedia "W" icon for these links:.
Because this book is on the World-Wide Web and there is no paper version of it, you must have a Web browser to read it. Be sure to use an up-to-date version of the browser software; otherwise, the pages will not display properly, and you may not be able to listen to the sound files. The best-known browsers that are usable across different platforms (in particular both computers running Windows and running Macintosh OS) are Netscape (use version 7 or later), Firefox (use version 1 or later), and Opera (use version 8 or later). For Windows users, another option is Internet Explorer. For Macintosh users, another option is Safari.
There are many links to sound files in the book. To see if your browser is set up properly for playing these files, click on this link: this is a recording. There are also some links to movie files. If your browser is set up for these, you should see a picture of a woman below and should be able to play the movie by clicking on the controls below it. (The movie shows the sign in American Sign Language meaning 'movie'. For this and other ASL movies in the book, I am indebted to the Communications Technology Laboratory for their Sign Language Browser.)
Terms usually appear highlighted like this. When important terms are introduced for the first time, they appear like this. When such terms appear later in the book, there is often a link back to their first appearance. All of these important terms are also listed in the glossary. Concepts are sometimes displayed like this. Emphasized words appear this way.
The book is divided into chapters and sections, with one webpage for each section. Sections are divided into subsections. Many subsections begin with an example and one or more questions to get you started thinking about the topic; these examples and questions appear in
boxes like this.
boxes like this.
Some sections may also contain less important portions that can be skipped. These appear in an indented chunk of text in a smaller font like this.
(something in the margin)
next to the part of the text that they refer to.
At the end of each chapter is a section containing problems on the material in the chapter. There is a link to the problems covering a given section at the bottom of that section page.
In the book, linguistic examples from languages other than English usually include a representation of the pronunciation of the word(s) and the meaning of the word(s). The meaning usually appears in the form of a gloss, that is, a word or brief phrase in English designed to give a general sense of what the expression means. Glosses appear between single quotes (' ').
For some linguistic examples the precise pronunciation is important; for others it isn't. When the pronunciation matters, the example is shown using phonetic symbols. A list of all of the symbols used in the book appears in this appendix. The symbols for sounds appear between slashes (//) or between angle brackets (); the difference between these two notational conventions will become clear in Chapter 3.
When precise pronunciation is not important, linguistic examples appear in italics. For languages like English that use the Roman alphabet, the standard orthography (spelling) is used. For languages that do not use the Roman alphabet (Chinese, Japanese, Amharic, and Inuktitut among our main group of nine languages), the examples are transliterated into the Roman alphabet. When an example consists of a complex word, a phrase, or a sentence in a language other than English, it will often appear in a standard three-line format. The first line is for the expression itself. In this line, words will sometimes be broken into constituent morphemes, that is, units of meaning, separated by hyphens (-). The second line is for the meanings of individuals words and morphemes. Here hyphens separate the meanings of morphemes that are separated by hyphens on the first line. Meanings for grammatical morphemes appear in small capitals, often abbreviated. When more than one word is used to indicate the meaning of a single morpheme, these words are joined by a colon (:). The third line is for a gloss for the whole expression, enclosed, as elsewhere, in single quotes. Here is a Spanish example.
In this example the word hablaban has been broken into three separate morphemes, and a meaning is given for each of these below the word in the second line. There are two aspects to the meaning of the third morpheme (-n), 3p (third person) and pl (plural), so these are joined by a colon. Don't worry if you don't understand what these morphemes are doing (or even what a morpheme is) at this point; all of this will be explained later.
The English used in this book is meant to be a relatively informal variety of standard written American English. As you will see in the section on what linguists study and the section on what linguists do not study, there is sometimes disagreement about what counts as standard usage. One advantage of writing a book on the Web is that I get to decide the conventions myself rather than being forced to conform to the standards imposed by an editor. Because this is a book about language and because particular usages can sometimes lead to negative, kneejerk responses, it may be worth mentioning two features of the English used in this book that do not count as standard for everybody. First, I will often use they, them, and their to refer to an unspecified singular person, as has been common in English speech and writing since at least the time of Shakespeare. An example is the sentence what language a child learns depends on what language they are exposed to. Second, I will avoid the word whom altogether. This word is very infrequent in any modern English spoken dialect and is not used in a consistent fashion in formal writing either.
Readers unfortunate enough to have been taught rigidly traditional English teachers will also notice split infinitives (to overwhelmingly reject this proposal), prepositions at the ends of clauses (as in the example with "singular they" above), and sentences beginning with and or but.
1.2 What we study
A science takes phenomena of one kind or another as its subject matter and attempts to describe and explain them objectively. Scientists gather particular kinds of data, analyze them, and create theories that account for the data. Here are some sciences and informal descriptions of the phenomena they are concerned with.
How would linguists and other language scientists go about objectively describing and explaining how language works? What kind of data would they examine? How would they analyze the data? What would it mean to "account for" the data with a theory? We will look at these questions in this section.
Social science and objectivity
You're a female American anthropologist specializing in the study of family structure and the roles that different family members play. You're studying a remote ethnic group, and you discover that the people believe that baby girls need to be brought up more harshly than boys to prepare them for a harsh life and that fathers have no particular role to play in the bringing up of their children. How might it be difficult for you to be objective in your study of these beliefs and behaviors?
Of the sciences listed in the table above, all but chemistry are concerned with human behavior. For all of these sciences, the work of the scientist is complicated by the attitudes that we have toward the behavior; that is, it may be difficult to be objective in our study. It is probably cultural anthropologists who face this difficulty most often. The anthropologist described in the box above, having grown up among highly educated people in modern US society, almost certainly believes that boys and girls should be treated equally and that fathers should play an important role in the raising of their children. But she must somehow put these beliefs aside in her work. Instead of labeling the people's practices as "wrong", she must attempt to see how they fit in with the other practices in the society, whether she notices consistent patterns in their behavior. If she is to make any value judgments about the behaviors, they should be based on whether the behaviors contribute to the stability of the society, not on her own feelings about them. Her job is to describe and explain the society, not to judge it.
As another example, consider the case of another anthropologist specializing in the scientific study of religious beliefs and practices. To do this objectively, he has to put aside whatever beliefs he himself has on questions such as the existence of God and life after death. There is no way he can pretend not to have such beliefs; he just has to try to keep them from getting in the way. Note that this is not a problem for chemists; in their work they don't have to worry about their prior attitudes toward particular chemical reactions.
When we attempt to study human phenomena such as religion or family roles or language scientifically, we do not deny the inevitability or even the value of our attitudes toward the phenomena. It's just that evaluating the phenomena and trying to convince people to behave or believe in a particular way is not the business of scientists (at least not the main business); it's somebody else's. A priest or minister has a very different purpose from an anthropologist specializing in the study of religion (although the priest or minister can probably benefit from the insights of anthropologists who study religion). A family counselor who advises a husband to allow his wife more freedom is not doing science (although the counselor can probably benefit from the insights of anthropologists who study the family).
It is important to come to grips with some of our preconceptions about language before we begin to approach language as the object of scientific study. The next section is about some of those preconceptions and where they come from.
Attitudes toward the speech of others
You overhear the following conversation.
What do you think of comments like this? Do speakers from some regions or speakers belonging to some social or ethnic groups tend to be lazier than others in their speech?
Differences between languages
Undeniably communities of people do tend to differ. To take an obvious example, food preparation is more important in some cultures than others; some cultures are famous the world over for their cuisine. For language, the differences are again obvious to anyone. It's not just that languages sound different. Some languages make distinctions — in sounds, in words, in grammar — that others don't; in fact most of this book is about just this topic. And people learning a second language often have trouble making the distinctions that aren't part of their first language. What we naturally notice, as speakers of a particular language, is what is "missing" in other languages and what kinds of mistakes second-language learners make in trying to speak our language. This may lead us, consciously or unconsciously, to think there is something deficient about the other language or even about the speakers of the other language. It is very difficult for us to see it from the other perspective, to see that we also fail to make distinctions that matter in the other language and have trouble making them when we try to learn that language.
For example, as speakers of English, we may be surprised to find that Japanese has no words corresponding to English a and the, words that are so basic to English we may almost take them for granted. And we may be struck by the errors that Japanese learners of English make in trying to master these words. Similarly, we are struck by the confusions Japanese learners may have in pronouncing English words with the sounds that we write with l and r, a distinction not made in Japanese. But these same Japanese speakers may be surprised when they first learn that English has only one word for 'you' (Japanese has at least six possibilities) and struck by the tendency of English-speaking learners of Japanese to always use the same word for 'you'. And they are similarly struck by the difficulty English-speaking learners of Japanese have with distinctions in vowel length and pitch change, distinctions that don't exist in English.
In fact there is no evidence that people in some cultures speak in sloppier or more elegant or more monotonous ways than people in other cultures. And while languages do differ in striking ways, these different features seem to balance each other out. As far as we know, all languages are equally expressive, equally logical. If you're not already convinced of this, I hope you will be after you have read this book.
Linguistic chauvinism and intolerance
So if impressions like those of the speakers in the box above have no basis in fact, where do they come from? There are three possibilities. First, these people may have been told by an authority, for example, an English teacher, that certain usages are just plain wrong. Clearly, the reasoning would go, anyone who knows this should not be using those forms. We'll return to this issue in the next section. The fact is that what is "wrong" is all relative. The girls quoted in the conversation in the box would almost certainly find it wrong to say it doesn't mean anything when speaking to each other. If they wrote it don't mean nothin' in a school essay, on the other hand, that would be another matter (though it would still not be reason to call them lazy; it would just be evidence that they had not learned the rules of the variety of English that is appropriate in school). Second, these people may have a stereotype concerning the group in question, and they may be transferring that stereotype to the speech of that group. Third, what they hear differs from the English they speak, and people may be quite intolerant when it comes to speech. Especially if they belong (or believe they belong) to a political, economic, or intellectual elite, their view may be something like the following: "the way I speak the language is the right way; any other way is wrong".
Whatever the reason for the impressions of A and B in the box, a linguist would respond to them by saying that the two girls were simply speaking a different dialect of English, a dialect with its own grammar differing from the grammar of the dialect of A and B.
Describing and explaining language
You're a linguist who wants to study the English spoken in the Caribbean island nation of Grenada. How do you go about this? What sort of data do you gather?
Data for research on language
Linguists and other language scientists, unlike A and B in the box in the last section, are interested in what people do, not what somebody thinks they should do. The English of the girls overheard by A and B is just as legitimate an object of study as the speech of any other group. To carry on their study, clearly researchers need to gather examples of language. There are two sorts of ways to get these.
Experiments on language
Rather than using words and sentences produced by speakers (or writers), linguists and (even more often) other language scientists sometimes gather other kinds of data. For example, they might record the acoustic properties of speech or the movements of the tongue, lips, and jaw during speech. Or they might present people with words, sentences, pictures, or movies (the stimulus) and see how they respond or how long it takes them to respond to them. The responses in experiments like these could involve
Given some data, a linguist or other language scientist has to do something with it. Much of this book will be concerned with what they might do, so what follows will be just an introduction to this topic. For every kind of research on language, there are two things to be considered: what aspect of language is being studied and what the research is supposed to accomplish (and how it does this). We'll start with the first and look at the second in the next section.
The content of research on language
Language, even a particular individual language, is far too complex a subject to be studied in its entirety by any one researcher. As already mentioned in the overview of the book, there is a higher-level distinction we can make concerning what research is supposed to be about, between the study of language as system and the study of language behavior. In either of these two cases, the language scientist is normally studying only some aspect of the phenomenon, one or more of four sorts of things about a language (or dialect): its sounds, its words, its grammar, and its use in context. A researcher interested in the sounds of the language (or dialect) might try to figure out what the basic sounds of the language are, how they combine to form words, or how speakers produce the sounds. Many researchers in this area believe that it is possible to study the sounds of a language more or less independently from the other aspects of the language.
A second area of research is the words of a language, usually thought of as organized in some sort of abstract dictionary, referred to as the lexicon. A researcher interested in words might study how speakers find words when they are formulating sentences or how abstract meanings build on simpler meanings (how is over in get over a problem related to over in jump over the puddle?).
Another possible kind of research would try to characterize what counts as a possible sentence in the language, that is, what's grammatical in the language. It is not as easy as it might seem to define this concept. We must be careful to avoid any bias on the speaker's part based on what they have heard from teachers of their language in school because what we care about is what people actually say, not what someone tells them they should say.
But we also cannot just treat any sentence that occurs as grammatical because people make speech errors. By "errors", we do not mean that they break rules that apply to dialects other than theirs (for example, by saying ain't or he don't). Instead we mean slips of the tongue, false starts, and hesitations. For example, the following example includes several speech errors.
People produce such "sentences" all the time, but they clearly also know that there is something wrong with them. That is, linguists probably do not want their descriptions of a language to include such sequences. So grammatical sentences are possible sentences that do not contain speech errors.
Degrees of grammaticality
But there is another complexity; grammaticality doesn't seem to be an all-or-none matter. That is, while some forms may be completely acceptable to all of the speakers of a given dialect all of the time and other forms may be completely unacceptable to all of the speakers all of the time, there may also be intermediate cases that are not so clear. For example, some English speakers use a chalk for a stick (or piece) of chalk; others would be less comfortable with this (though they would not find it as unacceptable as, say, a clay for a lump of clay). There may even be variation within a single speaker. For example, an English speaker may say for my wife and me on some occasions and for my wife and I on others.
Returning to the examples ridiculed by A and B in the box above, we see that by the definition of grammaticality that linguists work with, such sentences as this one may be perfectly grammatical for the girl who said it.
A complete description of the grammar of these girls (and the community of speakers that they belong to) would have to specify just what counts as a grammatical sentence (for example, sentence 3) and what doesn't (for example, he don't nothin' mean or he don't meant nothin'), possibly singling out areas of grammar where there is disagreement and variation among the speakers. This description would obviously have to say something about word order and about which forms can go with which other forms (meant is a perfectly good word in their dialect, but not following don't).
But some linguists are not satisfied with just describing the grammatical sentences because this says nothing about what those sentences are for. Instead these linguists are concerned with describing how meanings and functions of language relate to words and grammatical sentences (for more on this idea, see this section). So an account that includes sentence 3 above as a grammatical sentence (for some English speakers) doesn't help us understand how this sentence conveys information about some person familiar to the hearer (he) and about the speaker's belief about that person's intentions. This book follows this second position on what we should be describing, that is, that we should be saying how language accomplishes things for speakers and hearers.
Finally, those same linguists who are interested in how sentences convey meaning may also be interested in describing a fourth sort of aspect of the language, a sort of "correctness" that is different from grammaticality. A sentence can be grammatical and meaningful — that is, the words and grammatical patterns in the sentence can sound right and correctly describe some possible situation in the world — but the sentence can still be inappropriate. Consider the following sentence.
Of course not everyone who uses a language or dialect (see this section for the difference between dialects and languages) knows how to do so grammatically and appropriately in all situations. In particular, language learners have only imperfect knowledge of the language or dialect they are learning, and they can be expected to make errors. Children learning English as a first language may say "doos" for juice or me up when they want to be picked up. Teenaged speakers of English as a first language may still commit errors of appropriateness, using informal expressions such as bigtime in formal contexts. And adults learning English as a second language may say "diss" or "dees" for this or I make the homework for I am doing the homework. We consider these to be errors, but they are only errors from the perspective of the system defined by the behavior of the adult native speakers of the language or dialect that is being learned. Such examples can also be seen relative to the learner's own linguistic system, which has its own pronunciation, vocabulary, grammar, and patterns of usage. Researchers studying second-language learning often find it useful to treat the learner's knowledge of the second language as a sort of language in its own right, what they call "interlanguage".
This section has looked at the kinds of topics that interest linguists and other language scientists and the kinds of data that might look for to help them in their research. But we haven't thought much about what the outcome of the research is. What would it mean to describe or explain language? We'll look at these aspects of research on language in the next section.
1.3 How we study language
Accounts, generalizations, and theories
In addition to content, we can look at research on language from the perspective of what it is trying to accomplish and how it does that. First, we need to go back and consider again what science is all about. As we've seen, a scientist starts with a phenomenon of interest, gathers some data on the phenomenon, and attempts to come up with a description or explanation of it. This may take the form of a discussion in some language (such as English or Chinese), a set of equations, or an algorithm, that is, a precise description of a set of processes to be carried (by a person or a computer). Whatever form it takes, this result of the scientist's research may be referred to as an "account" or an "analysis" of the phenomenon. A scientific account is expected to include some sort of generalization about the phenomenon, that is, to go beyond simply listing the data.
How scientists describe and explain: generalizations and theories
To take a linguistic example, a researcher might be interested in describing the way present-tense verbs are negated in English dialects like the one referred to by A and B in the box in the last section:
The linguist could make the low-level generalization that for all but a few verbs in these dialects, the present tense negative is formed by putting don't before the base form of the verb (go, mean, eat, etc.). This generalization would apply only to the dialects being investigated.
Or the researcher could note that there is a distinction in the affirmative in these dialects that is not made in the negative, the same distinction that is made in standard English dialects:
Explaining languages, explaining language
Going beyond these English dialects, the linguist might then discover that something similar happens in many languages, for example, among the languages that this book focuses on, in Japanese and Amharic, making the generalization that languages tend to make more distinctions in the affirmative than the negative. The researcher could then go even further and try to place negation within a set of other forms in terms of how likely they are to favor distinctions; for example, the researcher might discover that fewer distinctions are made in verbs in subordinate clauses (for example, in the places they go) than in independent clauses (for example, in they go many places) in many languages. (Don't worry if you don't understand what "subordinate", "independent", or "clause" means.) Even more abstractly, the linguist could try to explain why there would be such tendencies across many languages. For example, they might propose that it is more difficult to produce and understand negative forms than affirmative forms and that this puts pressure on the users of the language, and hence on the language itself, to compensate by making some other aspect of the grammar simpler. This kind of proposal would be a theory that is designed to describe and explain a set of grammatical phenomena across many languages.
To summarize this section so far, we see that a scientist trying to understand language makes generalizations about data. These generalizations can apply only to one language, or they can apply to language in general. A linguist or other language scientist often works within the context of a particular theory of language. I'll have more to say about the role of theories in this section. But we've said nothing so far about what the goal of a description or explanation is. There are two kinds of possibilities, related to the two fundamental ways of looking at language that are mentioned in the overview of the book: focus on the product (sounds, words, sentences, etc.) and focus on the process (speaking, reading, understanding, etc.).
Product and process
Consider this line from the English comedy show "The Two Ronnies":
The joke, in case you missed it, is based on the ambiguity, that is, the multiple possible interpretations, of the verb left. The sentence could mean that the woman abandoned her husband in favor of the garbageman, or it could mean that she put him outside for the garbageman to carry away. It turns out that ambiguity is quite common in language and is the basis of puns such as this one. What sort of a problem does this present for a description or explanation of how language works? What aspects of language would an explanation of ambiguity need to refer to?
All instances of language are obviously the result of processes: speakers, writers, and signers (in the case of sign language) produce something that we call language and hearers, readers, and sign observers attempt to understand something that has been produced. Over a longer time scale, what a person knows about a language (that is, how to produce and understand it) changes; this is the process that we call learning, development, or acquisition. Over an even longer time scale, every language also changes; the English of today is not the same as the English of 1900. This slower process is called language change. Finally, over the longest time scale that is relevant for language, we know that at some point in the distant past, for example, 200,000 years ago, the ancestors of modern humans did not have anything like what we call language; the process that resulted in the kind of system we have now is called language evolution.
Most linguists choose to ignore all of these processes and focus instead on the products, the words, sentences, and entire discourses that are produced and understood by people at a particular point in historical time. In their research they attempt to describe and explain these products, to generalize about what must appear, what may appear, and what may not appear. Other language scientists, especially psycholinguists and computational linguists, focus instead on the processes themselves. In their research they attempt to describe and explain these processes, to generalize about what is going on during language behavior, language change, or language evolution, for example, when a speaker pronounces a word, when a child learns how to combine words into sentences, when a population of agents "invents" grammar in the process of evolving a communication system. Their accounts and theories often take the form of algorithms and are often implemented in the form of computer programs. They may be called processing accounts or computational models.
Production and comprehension
Second, processing accounts are directional. At its most basic, language processing is either production or comprehension. Even processing accounts that are concerned with the slower processes of learning or evolution are based on some idea of how production or comprehension takes place. I'll have more to say about production and comprehension later in this chapter. For now the important idea is that these processes occur in opposite directions.
In production, a speaker, writer, or signer starts with something to be communicated (perhaps to themselves) and then goes through a process that results finally in an instance of language (a spoken, written, or signed utterance of some kind). When I produce the sentence take the garbage out, I start with something that isn't language at all, something that doesn't include the words take and garbage but is more like my mental representation of some situation in the world, either one I'm experiencing (seeing the garbage piling up in the house) or one I'm imagining (seeing the hearer taking the garbage out). Then somehow I get from this thought to the utterance itself. In comprehension, a hearer, reader, or sign observer starts with an instance of language and then goes through a process that results in some sort of approximation to what the speaker, writer, or signer wanted to communicate. When I hear somebody say the sentence take the garbage out, I start with some sounds and from these eventually figure out what the speaker wants and what I'm supposed to do, and of course what I'm supposed to do isn't language at all; it's an action. The product-oriented perspective ignores this directionality, treating it as irrelevant to what makes language work.
So who is right? Does the product-oriented perspective or the process-oriented perspective give us more insight into how language works? To some extent, the answer depends on what we're after. If we want a way to describe languages in as efficient and understandable a way as possible, then it may be that we can confine ourselves to product-oriented research. The outcome of this research could be an archive of many languages in a form that would allow researchers from different fields and people who want to teach or learn the languages to consult the archive. Of course if our goal is to understand what people are doing when they are actually using language or if we want to write programs that allow computers to use language, then we will need to rely on process-oriented research.
But what if we are interested in the more abstract and theoretical question of why language is the way it is? Which perspective is the right one, or do we need both? There is a lot of controversy within the language sciences on this point, with one camp, associated especially with the famous linguist Noam Chomsky, claiming that we can learn what makes language special by studying products alone. The idea is that processing is something separate, something to be understood in its terms, but not something we need to refer to to understand how language works. The opposing camp, associated with the theories in linguistics known as cognitive linguistics and functionalist linguistics and with many language researchers in fields outside of linguistics proper, takes the view that the nature of language is intimately tied to the way it is used, that if we want to understand why language is the way it is, we need to refer to processing. As I discuss more in this section, this book belongs more in the second camp. I will assume that both product-oriented and process-oriented perspectives can help us understand how language works.
Let's return to the issue of ambiguity, illustrated in the box above, to see how thinking about process as well as product can help us figure out what is going on. For the sentence in question, a product-oriented approach might simply include the information that left (or leave) has at least two meanings, though of course it would have to be more precise about what is meant by "have two meanings". A process-oriented approach would look at the processing of the sentence from both directions. This would make clear that ambiguity is a "problem", that is, a potential challenge for a person or a computer, in the comprehension direction but not necessarily in the production direction. It is such a problem in fact that a very large body of research from the processing perspective has looked at what is called disambiguation. The problem is that when a word is ambiguous, a listener (reader, sign observer) has to figure out which meaning is intended (or, in the case of a joke like the one in the box, to see that the sentence has a possible interpretation for each of the meanings). It is easy to imagine how a person hearing a sentence like the one in the box would know that the word left has more than one meaning. What is hard to explain is how the person knows which meaning is the right one, or, in this case knows that both are right. Theories of disambiguation are designed to explain the process. We'll return to disambiguation later; the important point for now is that it only becomes an issue when we look at language from a processing perspective.
In this section and the last, we've seen that the business of linguists and other language scientists (and the subject of this book) is trying to describe and possibly explain particular languages, language in general, and how people use languages. But this is not the only way we might treat language. Some people who are not scientists and may not be particularly interested in the scientific study of language do talk about language as a part of their work. And one of their concerns may be deciding what people should and should not say and trying to enforce these decisions. This kind of work, some of its consequences, and how linguists sometimes comment on it are discussed in the next section.
The instructor of the course corrects you: "'He or she has', not 'they have'." How do you feel about being corrected in this way? Do you think the instructor was justified?
Just as it is normal to evaluate the religious beliefs and practices and the family behavior of particular individuals within our culture, it is also normal to evaluate the linguistic behavior of people. We may treat some speech and writing patterns as acceptable or unacceptable, superior or inferior, appropriate or inappropriate. Attitudes such as these must be based on a standard, some idea of what counts as desirable behavior. People who are concerned with defining and maintaining linguistic standards are prescribing (rather than describing) language. Usually they prescribe aspects of grammar, and in this role they are referred to as prescriptive grammarians. There are three kinds of questions we can ask of linguistic standards and language prescription.
Why a standard
One reason for a linguistic standard is that people within a community (for example, a nation) will be better able to understand one another if they agree on a set of words and a set of rules for pronunciation, spelling, and grammar. The process of defining the rules (and sometimes the set of words as well) is called standardization. Once a standard has been agreed on, it can be used in the media and taught in the schools.
Japan provides a good example of this process of standardization and promulgation of the standard. Children all over Japan, speaking widely divergent dialects of Japanese, are all taught to speak and write Standard Japanese in school, and radio and television announcers are all expected to be familiar with the standard vocabulary, pronunciation, and grammar. If these announcers were to speak in the southern dialect of Kagoshima or the northern dialect of Aomori, people in most of Japan would have difficulty understanding them. Something like this happens in the United States, though more informally. Children across the country are taught the same written standard vocabulary and grammar, though they may not be taught a single standard pronunciation, and radio and television announcers tend to speak in a single standard.
A related, though often unstated, goal of a standard may be to eliminate diversity, which may be perceived as a threat to national unity. Sometimes the diversity is reflected in different, though closely related, dialects, but sometimes it is reflected in completely different languages. It is one thing to teach a standard in the schools across a country. It is another to discourage people from speaking their local dialects or languages (for more on dialects and languages, see the next section). Sometimes the native speakers of the non-standard dialects or languages become willing participants in this process in their desire to be integrated into the society.
In Japan, for example, standardization sometimes had the effect of eliminating diversity. In the case of the Ryukyu Islands, the southernmost region of Japan, the dialects that are spoken there are so different from the standard that they may be considered a separate language (or languages). Today, following the repressive policies of school administrators in the first half of the twentieth century as well as the economic and social pressures for Ryukyuans to conform, this language is on the verge of extinction. This is a familiar phenomenon in many countries with minority populations. In the US and Canada, immigrant and American Indian students have often been prevented from speaking their home languages in schools. These educational policies were most notorious in the case of American Indian students, who were sometimes sent to schools which kept them separated from their communities and were punished for speaking in their native languages. These policies were one of the major factors leading to the extinction and near extinction of the majority of North American Indian languages. For more on the topic of language death and "endangered languages", that is, languages in danger of dying, see the websites of Terralingua and the Foundation for Endangered Languages, two organizations dedicated to linguistic diversity.
Standard languages and politics
Clearly language standardization is a political issue, and as such it is not really the business of linguists (though it is studied by sociologists interested in the social and political aspects of language). However, as linguists often become closely involved with the people whose languages they study, they may become advocates for these groups when their languages or their well-being are threatened because of the language policies of governments.
Other reasons are often stated for prescribing language, though these may mask the ultimate political reasons. These reasons include the supposed illogical or ambiguous nature of some constructions used by people, and linguists have also sometimes gotten involved in the debate in these cases because of their expertise. An example is the use of English they, them, and their to refer to a single person, usually of unknown or unspecified gender, as in the example in the box above. The idea is that since they is supposed to be plural, it should not be allowed to refer to one person. In cases like this, prescriptive grammarians are trying to actually improve the language or perhaps to preserve what they suppose to be an earlier, purer form of the language. The problem is that linguists and other language scientists, experts on the "logic" and the degree of ambiguity in language, can find nothing inherently wrong with any of these constructions, at least not with the usual examples of "bad" English constructions. In fact languages seem to have their own built-in "prescriptive" mechanism which weeds out whatever patterns don't work. It's a survival-of-the-fittest sort of arrangement, and it implies that those common patterns such as "singular they" (which has been around in the spoken and written language for at least 500 years) are quite fit linguistically.
Where standards come from
If we are to define a set of standards for a community, they have to come from somewhere. One possibility for the source is from other languages, and, strange though this may seem, standardizers have sometimes resorted to it. Take the "split infinitive", which prescriptive English grammarians sometimes argue against. An English infinitive is an expression consisting of to followed by a verb stem (don't worry now if you don't know what this is), for example, to go or to sing. So according to this rule, it is incorrect to say or write to boldly go or to not sing. The main argument seems to have been based on the fact that Latin and German and other familiar European languages cannot split their infinitives. But at least nowadays it seems just silly to try to force one language into the patterns of another.
It ain't about good and bad language; it's about class and education
If the standards come instead from among the speakers of the modern language, the question is which speakers. Language usage varies both with region and with social group, so if we want to standardize, we need to choose a region or group. As noted above, there are sometimes political considerations involved. In many countries, standard usage comes from the speech and writing of well-educated or upper-class people from a particular region, though the regional origin of the standard is not so clear in the case of the United States. For example, the current English standard prefers you aren't to you ain't because the latter form, although common in regional dialects all over the English-speaking world, is associated with uneducated speakers.
Actually this argument only holds today. The history is more complicated. Ain't was once used by upper-class speakers, but prescriptive grammarians who felt that it was somehow "lazy" or "illogical" succeeded in mostly obliterating it from upper-class speech in the English-speaking world. See this short history of ain't from the American Heritage Dictionary of the English Language (2000).
Another example is the pronunciation of pairs of words like pin/pen and since/sense as the same. This is a feature of the speech of a large region of the United States, including the South and some areas bordering on it, but it would not be considered a feature of standard American English.
How standards are enforced
Trying to teach grammar to native speakers
Although their efforts rarely seem to have an effect on children's speaking, in a limited way they may affect the children's written language. This is at least possible when the "rule" amounts to the prohibition of a particular form, for example, ain't. When it is more complicated, the prescribing may backfire, leading to behavior that is not what the prescribers had in mind. An interesting example is the use of "subjective" pronouns (I, he, we, etc.) vs. "objective" pronouns (me, him, us, etc.) in English. In an attempt to prevent usages such as him and me are friends, teachers have created a situation in which many speakers now say with he and I, instead of with him and me, not at all in accordance with the original intention of the teachers. This is an example of hypercorrection, which occurs when a prescriptive rule is applied in too many cases. For more on the issue of hypercorrection with English pronouns, see this interview with English professor Jack Lynch, which explains the usage but also takes a prescriptivist approach.
In summary, we've seen two sorts of usages that people try to enforce, usually attempting to prevent an "incorrect" form. In one situation, there is a form perceived as incorrect, such as ain't, that is characteristic of speakers of some dialects but not normally part of the speech or writing of speakers of the standard dialect. In another sort of situation, there is a form perceived as incorrect, such as "singular they", which is used by many or most of the speakers of the standard dialect as well as by speakers of other dialects. In this second sort of situation, there is often disagreement about what should count as the standard, and it is this kind of case where linguists sometimes become involved because of their interest and expertise in what people actually say or write. (My advice is not to get involved in one of these arguments. For some reason it often seems hard for people, at least for English speakers, to be rational about what counts as "correct" or "incorrect".)
I have focused on English. For a brief history of English "usage", that is, the concern for what is right and what is wrong in standard English, see this passage by E. Ward Gilman from the 1989 edition of Webster's Dictionary of English Usage. Not surprisingly, there are similar discussions and debates concerning usage in some other languages, though this sort of concern does not appear to be at all universal.
Other evaluation of language
So far we have looked at attempts to enforce linguistic standards on people. But people, especially composition and creative writing teachers, are also involved in evaluations of language based on other kinds of criteria. One is appropriateness, discussed in the last section. Appropriateness is one aspect of language that we continue to learn as adolescents and adults, and it is often explicitly taught. For example, a student might be corrected for using a word or phrase perceived as too casual in certain contexts, say, be into in the sense of 'be interested in' in an essay. Or an employee might be corrected by a colleague for calling his boss "dude".
Another important kind of evaluation concerns the effectiveness of language. A usage may be grammatical and appropriate to the context but still not accomplish the speaker's or writer's goal. For example, language may be needlessly ambiguous, as in the following example.
A composition teacher might criticize the student's use of it in this sentence; does it refer to the coconut or the screwdriver? Or a use of language may fail to accomplish a higher-level goal of the speaker or writer. Consider the following example at the beginning of an argument.
A teacher might find the writer's point trivial and obvious and suggest leaving it out.
These are all legitimate reasons for evaluating language of course. While linguists and other language scientists are not in the business of evaluating people's language, they can be of help by studying what appropriateness is, what makes an expression interpretable by a hearer or reader, and how the parts of a text relate to one another.
1.5 Dialects and languages
Idiolects and dialects
Two Americans are talking about a couple they have just met.
Two English people are talking about the same couple.
What's going on here? Who has the accent?
What I know about my language and how to use it is called my idiolect. It almost certainly varies in minor ways from the idiolects of all other speakers. But what is an idiolect? That is, what kinds of things do I know? In one sense, this whole book is an answer to that question, but we need to have a first cut at the answer here to help us get started.
I'll be much more careful later on about how each of these types of knowledge is described, but for now I'll say (informally) that my idiolect involves knowledge about vocabulary, pronunciation, grammar, and usage.
Of course no one is really interested in describing idiolects. Linguists and other language scientists study the speech of communities of people, not of individuals. More specifically, they study the knowledge of vocabulary, pronunciation, grammar, and usage that is shared by the members of a speech community. Because the members of the community agree on this knowledge, because it differs (at least in some ways) from the knowledge shared by other communities, and because it is mostly arbitrary, I will refer to the knowledge as linguistic conventions.
But what is a speech community? I will use this term to refer to any group of people that shares a set of linguistic conventions differing in some noticeable way from the conventions found elsewhere. You may know that in the United States people in some cities have some characteristic features in their pronunciation, although they are easily understood by people elsewhere in the United States. For example, people native to Pittsburgh are known for using you uns (or yinz) to mean 'you plural'. Here's an example from the (partly tongue-in-cheek) "Pittsburghese" website: if yinz wants served, raise your hands.. The number of conventions that distinguish Pittsburghers from other English speakers in the northeastern United States is actually pretty small, but because there is such a set of conventions, we can consider these people to be a speech community. The speech patterns, that is, conventions of vocabulary, pronunciation, grammar, and usage, of a speech community are called a dialect, so we can speak of a "Pittsburgh dialect".
Which dialect do you speak? There may be a number of possible answers.
What about larger communities? Pittsburghers share some speech conventions with speakers in other cities of the northeast and north midwest, for example, their pronunciation of the a in a word like hands, as in the example above (more on this pronunciation later on). And people in that larger region share some conventions with people in an even larger region encompassing speakers in most of the northern and western United States, for example, their pronunciation of the long English vowels (bite, beat, bait, boat, etc.). And people in that even larger region share many conventions with English speakers all over North America, including most of their grammar and usage conventions, as well as a number of pronunciation conventions, for example, the tendency to pronounce the words latter and ladder in roughly the same way.
This idea of larger and larger communities, each sharing fewer and fewer conventions, is an over-simplification in one sense. The fact is that the boundaries of the communities overlap in many ways. If we look at particular vocabulary, we may find a region with one boundary, whereas if we look at other vocabulary or at some pronunciation convention, we may find another boundary. For example, Pittsburghers tend to say pop (as opposed to soda or some other word) for carbonated drinks, and they share this convention with many speakers in the northern midwestern cities who also share their pronunciation of the vowel in hands, but not with speakers to the east of them, in New York City, for example, who share the pronunciation but not the word. (New Yorkers tend to say soda rather than pop.) Thus where we draw the boundaries around a dialect depends on which convention or set of conventions we're looking at. For more about soda vs. pop, see this interesting website.
Another way what I've said so far is an over-simplification is that there is great variation within any of these regions. Some of this variation has to do with the constant contact between dialects that is a fact of life in most communities. Some of the variation also has to do with the fact that people often know a range of ways to say things and they may sometimes avoid their local dialect in favor of a standard (see below) in certain situations.
Each of these shared sets of conventions, whether at the level of a small village, a subculture within a city, or a larger region, is a dialect. And a linguist can be interested in describing any level and any aspect of the dialect at any level (pronunciation, vocabulary, grammar, usage). The pronunciation associated with a dialect is called an accent.
What is a language? How would you tell someone (say, an alien with no knowledge of human culture) what English is, without using the word language?
Two dialects of one language or two separate languages?
But how do we decide when a collection of dialects is a language and not just another, more general dialect? As we've already seen, a dialect can also be a set of dialects (the North American English dialect consists of Southern dialect, New England dialect, Canadian dialect, etc.). What makes English a language and not just another very general dialect? What makes Canadian English a dialect of English and not a language in its own right?
The answer to this question is complicated. In fact there is no clear answer because the words dialect and language are used in different ways for different purposes. There are two completely different kinds of criteria related to the distinction between dialect and language, linguistic criteria and social or political criteria.
Given two overlapping sets of linguistic conventions associated with two different speech communities, for example, Mexican Spanish and Argentine Spanish, how do we decide whether they should count as two dialects or two separate languages? One criterion is the degree of overlap: how similar are the vocabulary, the pronunciation, the grammar, and the usage? Unfortunately there's no simple wat to measure this overlap, at least no way that researchers would agree on. One way to have a sense of the overlap, though, is mutual intelligibility, the extent to which speakers from the two or more speech communities can understand each other. Mutual intelligibility is also not easy to measure, and it is often based on the impressions of speakers and hearers, how much they understand when they encounter members of the other group or how long it takes them to get accustomed to the speech of the other group. We also need to establish some sort of intelligibility threshold; no two speakers can be expected to understand each other all of the time. So none of this is precise at all. The idea is simply that if two sets of linguistic conventions are similar enough so that their speakers can usually understand each other, then the two sets of conventions should count as dialects of the same language rather than separate languages. On these grounds, we call Mexican Spanish and Argentine Spanish dialects of the same language (Spanish) because speakers of these dialects normally have little trouble understanding each other.
To find out what should count as a separate language on grounds of mutual intelligibility, a good resource is Ethnologue, an online database of all of the world's known languages, 6,912 according to their current listing. The Ethnologue compilers attempt to use mutual intelligibility to decide what should count as a language. While English is listed as a single language, both German and Italian are listed as multiple languages. Each of these languages, for example, the variety of Italian called Sicilian, is usually referred to as a "dialect", but, according to the Ethnologue compilers, these are distinct enough to be considered separate languages. Again, the criterion of mutual intelligibility is a rough one, and some of Ethnologue's claims are controversial.
Social and political criteria
Mutually intelligible "languages"
At the other extreme are examples like the languages spoken in the northern European countries Sweden, Norway, Denmark, and Iceland. These "languages" are all related to one another, and speakers from some pairs of countries within these have little difficulty understanding one another when they are speaking the standard dialects of their languages, despite the obvious differences, especially in pronunciation. Thus on grounds of mutual intelligibility, we might consider some of these "languages" to be dialects of a single language. But Icelandic, Swedish, Danish, and Norwegian are official "languages" of separate countries, and there are separate spelling conventions for some of the sounds in the languages.
Actually the situation is even more complex than this because Norway has two official dialects, and a fifth related language, Faroese, is spoken in the Faroe Islands, which are administered by Denmark.
So for mainly political reasons, they are considered separate languages rather than dialects of a single language.
To summarize, the line between dialects of one language and separate languages is somewhat arbitrary. However, wherever we draw the line, three points should be clear.
The following appears on the website of a person who spent some time in Pittsburgh: "probably relating to the rest of Pittsburgh's terrible dialect, which I, fortunately, did not pick up". Why would some dialects be thought of as "terrible"?
Some dialects within a language may be singled out for special status. When we're dealing with a political unit, such as a nation, in which related dialects are spoken by most people, one dialect is often treated as the standard dialect. You know something about this already from the last section of the book. The standard dialect is often the only dialect that is written, and it is the one that is taught in schools and (with some exceptions) used in the media. Thus in Germany, Austria, and the German-speaking part of Switzerland, it is Standard German that is taught in the schools and used in broadcasting, even though most people in this region are not native speakers of the Standard German dialect. This means that most people in the German-speaking countries end up bidialectal. The same situation holds in Japan, where it is Standard Japanese, based on Tokyo dialect, that is taught in the schools and used in the media. Note how this makes it possible to speak of a German or Japanese speech community, even when the native dialects of people in these communities are very different from one another, because all educated speakers in these communities share the standard dialect, often as a second dialect.
Dialects of English in the US and England
In the United States, the situation is somewhat simpler than in Germany or Japan because the differences among most of the dialects are not nearly as great; native speakers of English in the United States have little trouble understanding each other. (An important exception is African-American Vernacular English (AAVE), spoken mainly by many African-Americans.) As in Germany and Japan, we have an (informal) standard dialect, for vocabulary, grammar, and usage, if not for pronunciation. Thus children in Pittsburgh learn in school to write sentences like the school needs to be renovated rather than the school needs renovated, which would be grammatical in their local dialect. Americans tend to be relatively tolerant of differences in accent, however. Teachers in schools throughout the country teach the standard grammar but use their own local pronunciation. If we have a standard accent, it is the one people associate with television announcers, the accent characteristic of much of the Midwest and the West. This accent is called General American; I will have more to say about it later.
The situation in England is similar to that in the United States, and the standard vocabulary, grammar, and usage that children learn to write in English schools are very similar to the American standard. However, in England, there is a stronger idea of a standard accent than in the United States and more pressure for children to learn this accent if it differs from their home accent. This accent is referred to as Received Pronunciation (RP); it is based on the speech of educated speakers in southern England. (Note that RP is standard English English pronunciation, not British English; in Scotland, there is a quite different standard accent.) I'll have more to say about RP and how it differs from General American and other English accents in this section.
"Just" a "dialect" or a full-blown "language"
The existence of a single standard dialect among a set of non-standard dialects has important social implications. The non-standard dialects have less prestige, and their use may be discouraged in formal situations, not just situations in which writing is called for. Sometimes, as in the Ryukyu Islands in Japan or in some regions of France and Spain, this leads to the decline and possible death of the non-standard "dialects" (which would be considered languages by the mutual intelligibility criterion). In other situations, speakers of non-standard dialects retain pride in their local speech patterns, while recognizing that they are not appropriate in certain situations. Finally, this pride, along with other cultural differences separating the speakers of the non-standard dialect from the speakers of other dialects (non-standard or standard), may lead to pressure to have non-standard dialects given official status, especially if they differ significantly from the standard. At this point the words dialect and language become politically charged terms because the supporters of official status for the non-standard dialect may feel the need to argue that it is not "just" a dialect of the larger language but rather a language in its own right. This has happened in the United States with AAVE (here is an essay on this topic by the sociolinguist John Rickford) and in Europe with many languages that are normally considered "dialects" of other languages (this website includes many of them as well as links to other sites concerned with the "minority language" question in Europe and elsewhere).
Why some languages resemble each other
We've seen how as we extend the boundaries of speech communities, we get fewer and fewer shared conventions. When we reach the level of a language such as English, Spanish, or Mandarin Chinese, we have a speech community which shares a set of conventions (in some cases a standard dialect) which allows people in the community to communicate with one another despite dialect differences. But we can go beyond a language. So for English, we could extend the boundaries to include the Netherlands, Germany, Scandinavia, and some other regions in western Europe. We'd now find a much smaller set of shared conventions. All speakers in this large "community", for example, share a word meaning 'all' which is similar in pronunciation to the English word all. But there would be no reason to call this set of conventions a "language" since the speakers obviously do not understand each other and do not belong to a single political unit with a single standard dialect. Instead we refer to this set of conventions, or set of languages, as a language family, in this case, the Germanic languages. The members of a language family resemble each other because they are genetically related; that is, historically they derived from a common ancestor language. (Note that this use of the word genetic differs somewhat from its use in biology; the speakers of Germanic languages are not necessarily genetically closer to one another than they are to the speakers of other languages.) The ancestor of the modern Germanic languages was not a written language, so we can only infer what it was like.
In most cases we can go even further back; the ancestor languages of two or more families themselves may have had a common ancestor language. Thus the modern Romance languages, including Spanish, French, Italian, Portuguese, Catalan, and Romanian; the modern Germanic languages; and many other languages spoken today in Europe, the Middle East, and South Asia, apparently descended from a much older (and also unwritten) language. This means we can group all of these languages into a single family, in this case the one we call Indo-European. Sometimes, to distinguish the lower from the higher levels within a family tree of languages, we use "language family" only for the largest grouping (for example, Indo-European) and "branch" to refer to groupings within this (for example, Germanic and Romance). Note that there may be many intermediate levels in the family tree of languages. Within Germanic, for example, there is North, including the Scandinavian languages, and West, including English, Dutch, and German.
Note also that languages may resemble each other in one way or another for reasons other than a genetic relationship. The main non-genetic source of similarity is language contact; when the speech communities for two language are in close cultural contact, their languages often influence one another. So modern Japanese vocabulary includes thousands of words borrowed from Chinese and uses the Chinese writing system (as well as writing systems specific to Japanese). But, except in the sense that all human languages may be ultimately related to one another, there is no evidence that Japanese is genetically related to Chinese. A more complicated situation occurred in Western Asia with the complicated cultural influences among people speaking Arabic, Persian, and Turkish. These three languages belong to separate language families (Afro-Asiatic, Indo-European, and Altaic, respectively), which are either unrelated to one another or only very distantly related, but Turkish and Persian have borrowed many words from Arabic, Turkish has also borrowed many words from Persian, and Persian borrowed its writing system from Arabic.
1.6 A stand and two themes
Linguistics is a relatively young science, and psycholinguistics and computational linguistics are even younger, so it's not surprising that these fields are still fraught with controversy. Though there is general agreement on some core topics, some quite basic issues are still up in the air. It is not the place of an introductory textbook to go into all of the controversies; I will only try to do this for one topic, language learning, perhaps the most hotly debated of all. But it is impossible to discuss any topic related to language without taking some sort of stand on the questions that divide language scientists. All linguistics textbooks take such a stand, though, sadly, few of them tell you that they are doing this. The stand taken in this book is that language and language behavior are not phenomena separate from the rest of human perceiving, acting, and reasoning; that we can only understand how language works by understanding how it fits into the rest of human behavior. Another way to think of this position is that we will be treating the language sciences as cognitive sciences.
This position is related to two of the ways of looking at and studying language that we've already discussed. As noted in the overview of the book, we can study language as a system independent of the people who use it, or we can study language behavior, focusing on use. As noted in the section on how language is studied, we can be concerned with product or process. The stand that this book takes implies that it cannot be enough to study language as a system in its own right (though it might be useful some of the time to do this) or to study product rather than process (though again it may be useful some of the time). It also implies that the border between language and non-language is not necessarily a clear one and not something we should spend much time worrying about. So for every topic covered in this book, I will try to look at it both ways, to see how the system/product perspective and behavior/process perspective both contribute to an understanding of how language works.
I didn't invent this position of course. A number of other people have defended one or another aspect of it. On the need to consider cognitive processes outside of language in order to understand language, cognitive linguists like Ronald Langacker and George Lakoff have made this case in many of their writings, in which they invoke general psychological notions such as attention, categorization, and memory. On the need to treat language as process, an very convincing argument was made in 1983 by the computer scientist Terry Winograd in his important book Language as a Cognitive Process (1983).
In the rest of this section, I discuss two themes that are consistent with this stand and that will guide the rest of the book.
Meaning, function, and form
The Main Theme of the book is that language associates meaning and function with form and that understanding how language works requires that we focus on these associations. All three of these terms will become clearer later in this book. For now we will just try to get the gist of what's intended by them.
Linguistic form concerns the way a linguistic expression sounds (for spoken language) or looks (for written or signed language) or how a linguistic expression is produced. There are two ways of looking at form. The usual way within linguistics is to think of form as a sequence of elements. For written language this is a natural way to treat form because written sentences are sequences of elements (characters). For spoken or signed language, it is less so. Treating a stretch of spoken or signed language as a sequence of elements means in effect focusing on the written transcription of the language in terms of units of sound or primitive gestures of some sort. This approach tends to play down the role of time, to look at a stretch of language as a static object. An alternative and more radical way of looking at form (at least for spoken or signed language) is to treat a stretch of language as something inherently continuous and dynamic, something that cannot (or should not) be transcribed as a written sequence of elements. In this book, I will usually discuss linguistic form in the first way, mostly because this is the way it is usually done and because it seems to simplify the analysis. But I will also insert periodic warnings about the biases that are built into this approach. One particular bias we should be aware of is one based on literacy; that is, as skilled users of alphabets, we may tend to think of spoken and signed language in this way (for an argument of this sort, see this paper by linguist Robert F. Port).
Concepts without words
Whatever we mean by concepts, it is clear that not all of them are associated with words. Words are linguistic; concepts need not be. We all have concepts that we have no words or grammatical patterns for. For example, one concept I have is the little depressed region bounded by two vertical ridges that is found between the nose and the middle of the upper lip. I have no idea what to call that place (though I'm pretty sure it has a name in at least some languages); that concept does not represent the meaning of a word for me.
Another point to note here is that meanings do not need to be seen as thing that are "out there" in the world. Since the stand taken in this book is that language is basically a cognitive phenomenon, this would be a strange way to think about meaning. Instead I will be treating meanings as things that are "in here", cognitive entities realized in our brains, depending on our interpretation of what's out there and including imagined entities that aren't "out there" at all (though they are inevitably based on things that are).
I will use function in two senses, first, for the uses that people put language to and second, for the uses that particular words or patterns have within stretches of language. It is the former sense that concerns us here. People use language to refer, to assert, to command, to convince, to get information, to entertain, to deceive, and much more, and these uses of language obviously have something to do with the forms that they choose. (If you want somebody to lend you their computer, you don't say you will lend me your computer, you say something like could I borrow your computer?.)
Many people who study language make a distinction between function in this sense (part of what they call "pragmatics") and meaning (what they may call "semantics"). I would argue instead that the difference is a matter of degree, but that there's not much point in worrying about where the line between pragmatics and semantics is.
A second theme of the book is that language is the way it is in part because of constraints coming from the nature of human biology and human cognition. A constraint is a kind of limitation on what is possible. Consider constraints that come from the nature of the human body itself. For spoken language, the physical and physiological properties of the vocal tract constrain what can be produced, and the auditory system constrains what can be perceived. For sign language, the physical and physiological properties of the hands, arms, upper body, and face constrain what can be produced, and the visual system constrains what can be perceived.
Just as important are cognitive constraints, constraints arising from the nature of the human mind. First, there are limitations on human memory. Cognitive scientists divide human memory into short-term and long-term memory. Short-term memory is used to temporarily store the information that is needed as we interpret the world around us. When we are faced with a complicated visual scene, we can't look at everything at once, and we have to scan the scene in order to come up with some kind of interpretation. But we need to temporarily store some sort of record of what we've already seen in the scene as we do this. That information is kept in short-term memory. Likewise, when we are listening to a sentence, we need to remember the parts that we've already heard as we are trying figure out the meaning of the whole. Short-term memory is used for this as well. The same applies to reading and to the visual processing of sign language. Cognitive science research has shown that our short-term memory has a very restricted capacity, and this places strong constraints on how sentences and discourses are organized.
We use our long-term memory to store information that we learn through experience. Information may remain in long-term memory for a very long time, even indefinitely, and human long-term memory is very large, larger than the memory of any current computer. But long-term memory is still limited. We obviously do not have an infinite memory capacity, and just because we have stored an item in memory doesn't mean we can retrieve it when we need it. We will see later how the finiteness of long-term memory matters for how language works.
Speaker orientation and Hearer orientation
Finally there are constraints that are specific to the two ways in which language is processed, language production, that is, speaking, writing, and signing, and language comprehension, that is, listening, reading, and interpreting linguistic signs. To simplify matters, I will be referring to a language producer as the Speaker, even though writers and signers as well as speakers proper are intended. And I will be referring to a language comprehender as the Hearer, even though readers and sign interpreters as well as hearers proper are included. (To remind ourselves of these distinctions, I will capitalize Speaker and Hearer when they are used in this way.)
For the Speaker, the main constraint is that the production of linguistic forms be easy. It is easier for the Speaker to make relatively few distinctions because the Speaker has to remember what the distinctions are and to make the extra effort to keep things distinct. For example, maintaining the agreement between subject and verb in the present tense in English (the girl sings, the girls sing) requires an effort on the part of the Speaker. What is easy for the Speaker depends in turn on constraints from the body. For example, large movements of the tongue tip are more difficult to execute than short movements, so short movements should be preferred from the Speaker's perspective.
For the Hearer, the main constraint is that linguistic forms that need to be distinguished can be easily distinguished in comprehension. This constraint also depends on the body, specifically the parts of the nervous system that are responsible for hearing (for spoken language) and vision (for written and signed language). For example, if a language contains a large number of homophones, that is, different words which sound the same (such as two, too, and to in English), this may put a burden on the Hearer.
Speaker-oriented constraints and Hearer-oriented constraints often oppose each other; what simplifies things for the Speaker (for example, not making many distinctions) complicates things for the Hearer. We will see many examples in the book of these opposing tendencies.
The opposition of Speaker orientation and Hearer orientation is particularly clear as languages change. Languages change for a variety of reasons — contact with other languages, imperfect learning by children, random fluctuation — but it appears that all languages are always changing. Most of the changes can be seen as either Speaker-oriented or Hearer-oriented. For example, the grammar of a language may become simplified as some suffixes are dropped, a change that seems to result in less work for the Speaker. But the two kinds of trends always balance each other out in the end, and the simplification of the grammar in one way will probably be compensated for by an increase in complexity (from the perspective of the Speaker) somewhere else in the grammar of the language. Otherwise language would fail as a communicative device. These built-in pressures in favor of the Speaker and the Hearer apparently prevent the world's languages from moving in some general overall direction. That is, at least in recent history, it does not appear that languages have generally been getting simpler or more complicated (in any sense of these words) as they evolve.
1.7 Why study language
What good is the scientific study of language? Why does anyone do it? Why should you care about it? These are the sorts of questions you have a right to ask about any university course. The answer to the last question depends a lot, of course, on how you happened to end up in a course using this book in the first place and on what your interests and long-term goals are. Language is a part of everyone's life, but it is more central to some people than to others. But I happen to believe that a scientific look at language should be a part of the basic curriculum, like mathematics and history are.
Many of you have already studied one or more languages other than your first, and more of you will later on. A few of you may teach a foreign language. In either case, you are not likely to find the learning process an easy one. Some of the difficulties faced by second-language learners have to do with differences between their first and second languages, differences in pronunciation, vocabulary, grammar, and usage. You do not have to know linguistics to learn a second language; after all, people all over the world who have never heard of linguistics do this successfully all the time. However, knowing what pronunciation, vocabulary, grammar, and usage are; how they can differ between languages; and how people seem to learn them (as first- or second-language learners) can help you be aware of and understand your problems and possibly correct some of them. A second-language teacher needs to be able to focus on problem areas, for example, by giving lots of practice or by simplifying other aspects of the language being learned. It is difficult, if not impossible, both to understand the source of the problem and to come up with ways of addressing it without understanding the nature of the material being learned, that is, what linguists and other language scientists study.
Some students, when beginning a linguistics course, believe it will help them improve their knowledge of their first language. In fact this is not something you can expect from a linguistics course. You already know the great majority of the words, the grammatical patterns, and the usage conventions that you will need to survive. Of course you can improve; you can learn new words and expressions, become more proficient with the grammatical patterns that are part of formal language, and get better at using language to accomplish your goals. Most of the improvement should come naturally as you are exposed to the complex language of academia and the workplace. But you will also face evaluations of your language by other people — teachers, colleagues, supervisors, even family and friends — throughout your life. You will be in a better position to make use of this criticism and advice if you understand what sort of problem is involved (if there really is a problem) and how it fits into the larger scheme of things. This is where linguistics can help.
Another way in which most people encounter first-language learning is in raising children. Of course you don't have to know linguistics to know how to "teach" a baby a language; babies aren't really taught language anyway. But knowing what it is that babies learn when they learn a language can make the process more enjoyable. You'll be able to better appreciate what an amazing process learning a language is, why so many people are fascinated with how this process happens, how your baby does the same kinds of things that others do and, at the same time, how your baby's learning steps are unique.
You may end up in a job that involves first-language learning more directly, as a teacher of your first language to native speakers. In this capacity, part of your job will probably be to make sure your students are competent in the standard dialect of the language. As we saw in the section on what linguists study and the section on prescription, sorting out what belongs to the standard is not a trivial matter, and a knowledge of what linguistic conventions are seems essential. As I've already said, there is a lot of confusion and controversy about what should be emphasized or even taught at all. You will also need to teach your students what counts as appropriate and effective language. Again linguistics and other language sciences can help; some language scientists devote their efforts to figuring out what makes different expressions appropriate in different situations, while others are concerned with how words and expressions are interpreted by hearers and readers.
Or you might work as a speech therapist, dealing with people with speech disabilities of one kind or another. Here the relevance of the scientific study of language is obvious; you first need to know what the norms of a language are before you can hope to address the ways in which your clients or patients deviate from these norms.
Linguistics and tolerance
What we can learn about the human mind by studying how languages differ from one another is more controversial. Linguists and other cognitive scientists disagree on how deeply the nature of a person's first language influences how the person thinks and views the world around them. I'll have more to say on this topic at various points throughout the book. But language is such an important part of our lives that learning about the languages of other people, including how those languages differ from ours, is in a very real sense learning about those people. As with any other aspect of culture, lack of knowledge can lead to intolerance. It is easy to believe that other languages are inferior in one way or another to ours, to think that some languages, especially the languages of relatively small ethnic groups in the Third World, are more primitive than others. So the realization that a language like Tzeltal, Lingala, Amharic, or Inuktitut has a set of grammatical categories and communicative options not even found in English is an eye-opening experience. Looking more closely at languages, and in particular at languages that might seem exotic to us, can make us more tolerant.
Finally, language is what we use to influence the beliefs of one another. This happens in everyday conversation, as we argue about who forgot to put the mayonnaise jar back in the refrigerator or what the results of the latest election mean. It happens in advertising, as companies do their best to get us to buy their toothpaste, shoes, and cars. It happens in education, as educators provide us with what they say are truths and try to convince us that knowing how to speak Spanish or how to do a t-test will be useful to us later in life. It happens in politics, as politicians and political activists try to get us to vote for them or to support their program. None of this is new, but the enormous quantity of information that is now available to most of us is new, and most of this information is designed to change our beliefs in some way. Obviously an educated person needs to be able to navigate their way through all of this, to sort out the nonsense, to see how bias and ideology are behind what is being claimed, to be a critical reader and listener, to make informed decisions. I don't believe that any of this is possible without understanding the role that language plays in knowledge, belief, and persuasion.
Note: This problem set assumes some basic knowledge of English dialects, so if you are not a native speaker of English, you will probably want to collaborate with somehow who is.
Each of the following includes two ways to say (at least roughly) the same thing. For at least some people, the first way (A) could be seen as "wrong" in some sense. We saw in these two sections of the book that there are various ways in which a particular linguistic form can be thought of as "wrong" (though not necessarily by linguists).
For each of the following, decide which type of "mistake" is involved in A (as compared to B), and explain your answer in a sentence or two. There may not be a single right answer for some of the problems.
You are a linguist studying the grammar of a particular dialect of English. You have recorded a stretch of speech between two men who are speakers of the dialect, and now you are taking examples from the recording to use as data. Included in the recording is the following statement by one of the speakers.
You are trying to decide whether your account of the grammar of this dialect should include this sentence. Unfortunately you no longer have access to the two speakers, so you can't ask them questions. For each of the following, say whether it would be a relevant thing for you to do in order to help make your decision. Explain each answer.
In this section, we saw that ambiguity is an example of a phenomenon in language that can best be appreciated from the perspective of comprehension. Consider the English words that are written bank, meaning roughly 'side of a river,' 'kind of financial institution,' and 'tilt while making a turn' (there are other meanings as well). Because all of these are written (and pronounced) the same, a reader (or listener) has to disambiguate the form when it appears. But this can be done in different ways, using different kinds of information. For each of the following sentences, say what kind of information a reader could use to disambiguate the word bank, and rank the three for how difficult they would probably be for a computer program. Imagine that the person or program knows only the three meanings of bank given above.
1.8.3 Two themes
For each of the following facts about human language, say whether it derives from Speaker (or Learner) orientation or Hearer orientation, that is, whether it makes things easier for the Speaker, the Hearer, or the Learner. Explain your answer in a sentence. There may be more than one possible answer.