Translators work with both speech and the written word. In the former case, the results are needed within seconds for communication to proceed; in the latter case, the results may take hours or even years. While translation of the written word has the luxury of more time, the task is accordingly more difficult than the translation of speech. A conversation, say, between the leaders of two countries will tend to touch on a limited number of subjects, with individual utterances being of generally short duration. A written work, on the other hand, may deal with a great variety of subjects, and the structure of the language may be much more complex. A reader has the opportunity to think more about what has been written, and thus has the opportunity for exploring multiple levels of meaning.
Regardless of the type of material to be translated, a translator has the formidable task of "living" in two (or more) worlds. In other words, the translator must maintain mental models of the cultures involved in the message, keeping in mind the pertinent conventions and experiences. A successful translator is one who can identify where these conventions and experiences overlap at different levels of meaning, and employ these areas of overlap to enhance the receiver's understanding of the message. One reason early attempts at machine-based translation failed so spectacularly is that they largely ignored the issue of meaning and concentrated instead on words.
As in other fields of human endeavor, most translators have the advantage of building on the work of others who have preceded them. Specifically, they do not have to learn about a language in the process of learning the language. Pioneering translators face just that task, however, where they must learn a great deal about the language and the people before a written work can be translated. This paper seeks to explain the work involved in such a translation project. It will not explain the need for such work, which is significant, nor will it seek to explain why individuals might be motivated to such work. These topics are properly the subject of other papers.
The audience for this paper consists of people who would like to understand what is expected of the translator in pioneering work. Although the subject areas are presented sequentially, the reader should understand that the translator may conduct many activities simultaneously. Not all tasks will involve equal effort, and the effort needed in a given area will depend both on the translator's ability and the difficulty of the task at hand. This paper will have served its purpose if the reader gains some appreciation for the work of translation and the people who undertake it.
Many people have had some experience in learning another language, either from school or from living in a foreign country. A course of instruction typically includes mastering the grammar and acquiring a working vocabulary. When living in a country where the language is spoken, the learning experience is often organized on the basis of what a person needs to know to carry out everyday activities. Either case presumes the existence of a body of knowledge about the language, which is something seldom available to the pioneering translator.
In a new culture, the pioneering translator is reduced to the level of a child. How does a child learn to speak? The child learns some individual words, and then starts putting them together. Having mastered a few words, children continue the learning process by asking lots of questions. Older speakers will generally cooperate by answering the endless questions, and providing correction where necessary.
A translator builds a vocabulary in the same way. After gaining the cooperation of one or more individuals, the translator has "helpers" to identify simple objects such as sticks, leaves, and rocks. This technique works for concrete objects, but fails for intangibles such as feelings. The translator may also have a hard time distinguishing related words such as stick, branch, and twig. To further complicate matters, the translator's native alphabet is probably insufficient for recording new words, especially if the new language contains unfamiliar sounds.
Consider the case of an English speaker trying to record Spanish words. Upon hearing the Spanish word for "tomorrow," the English speaker might try to write something like "manyana." By using the two letters "ny," the English speaker is compensating for the lack of a single letter to represent a Spanish sound. The results of this process may be seen in the English word "canyon," which is actually a loan word from Spanish. Beginning readers of English must be taught, however, that the word is pronounced "can-yon" and not "cany-on."
Linguists deal with the problem of recording sounds by using a tool called the International Phonetic Alphabet. Unlike other alphabets, this alphabet has characters to represent every sound that humans can make. The translator begins by writing words using this alphabet, listening carefully to how the individual sounds are produced. Since repetition of words is tedious for any speaker, translators often make recordings so that they can verify that they have judged the production of sounds correctly. Even a correct record of sounds, however, will not help to resolve problems such as words that sound the same but have different meanings.
Learning how speech sounds are made is part of the basic training for any linguist. Such training enables the linguist both to identify sounds that other people make and to produce the sounds that make up a given language. In analyzing a new language, a person will typically find that it consists of "familiar" and "unfamiliar" sounds. The familiar sounds may actually prove more troublesome, since the similarity may be only superficial. In such cases, supposed familiarity may lead to wrong conclusions. In learning a new language, one must be prepared to "unlearn" certain things acquired from other languages.
Although phonetics helps to understand what sounds are used in a language, it does not indicate the significance of those sounds. That is, native speakers may not make a distinction between different sounds. A simple example is the English word "paper," in which the two "p" consonants differ phonetically. Since English speakers do not perceive a difference in the two consonants, English readers have no problem with the two sounds being represented by the same symbol. Learning the sounds of a language is not sufficient; one must also learn which sounds that native speakers distinguish.
Phonological analysis is further complicated by intonation and tone. The former refers to the natural variation of pitch in conversation; the latter refers to changes in pitch that affect the meaning of specific words. Spoken with different intonation, the English word "really" can indicate surprise or insistence, while the basic meaning is unchanged. As with sounds, the learner must be careful about assumptions, since the significance of intonation in a new language is likely to differ from previous experience. Incorrect intonation can easily be a source of misunderstanding.
In contrast to intonation, not all languages exhibit tone. Tone works by changing the meaning of individual words so that utterances that are phonetically equivalent may nonetheless express different things. The distinction is somewhat akin to the changes that occur in English when the emphasis is placed on different syllables in a word. For example, the word "record" is a verb when the emphasis is on the latter syllable, and a noun when the emphasis is on the former. In languages that use tone, incorrect usage can make a speaker's utterance incomprehensible.
Readers are typically hard pressed to rememberwhat it was like before they could read. In teaching children to read, we generally concentrate on symbol recognition in conjunction with sounds. From there, children learn to combine letters to make new sounds, and they go on to form syllables and words. All of this presumes the existence of an alphabet, which frequently does not exist in a pioneering work. Without an alphabet, people cannot record things in their own language, and they are thus limited to oral transmission of the important aspects of their culture.
People from Western countries tend to think of the Roman ('abc') alphabet as the predominant alphabet used in the world. The truth is that over two-thirds of the world's population uses non-Roman scripts; Arabic and Chinese are just two examples. When refrring to writing systems, the term "script" is more accurate because some scripts are not properly alphabets, being composed of symbols representing syllables instead of sounds. Although non-Roman scripts have traditionally been written by hand, many scripts can now be composed on computers.
Developing a new script is an iterative process. First, a thorough linguistic analysis is necessary to identify the types of consonants and vowels in the language. Second, external influences have to be considered; for example, some speakers may have an affinity for an existing script because they first learned to read in another language. When tests begin, it may become clear that some essential distinctions between sounds have been missed because different sounds are being written the same way. On the other hand, readers may become confused because different characters are being used to represent the same sound.
Tone further complicates the development of an orthography. Without some indication of tone, different words can appear to be the same word. Although the meaning may sometimes be discernable from the context, tone markings are usually essential to ensure the reader's comprehension. A number of schemes have been used to represent tone; in some cases accent marks have proved sufficient; in other cases, superscripted numbers have been used. The best method for a given language will need to be determined by tests with readers.
Despite having attended "grammar school," most people learn relatively little about grammar. Rather, most of what they do learn about grammar happens in the first few years of their lives. These early years place no emphasis on terminology and they stand in sharp contrast to the sort of instruction a person receives when learning a foreign language in a formal setting. There, the language learner is confronted with ideas of direct and indirect objects, transitive and intransitive verbs, and the various rules that govern the construction of sentences.
Grammar presents a particular challenge for the pioneering translator. On the plus side, all languages exhibit grammatical regularity. On the minus side, the speakers of a language are frequently unable to explain the rules which govern their language. The person trying to discover the "rules" of a new language is thus cast upon the task of gathering enough samples of similar utterances so that patterns may become evident. These patterns must then be used to confirm or deny assumptions made about the phonology of the language, which may in turn affect the orthography.
Much of grammatical analysis is concerned with populating categories. All languages have at least nouns and verbs; to these various languages add adjectives, adverbs, and other categories. Within these categories are further distinctions: nouns, for example, may be separated into count nouns ("grain"), mass nouns ("sand"), and proper nouns ("Long Beach"). The different types of nouns function in different ways; in English, we do not say "much grain" (when referring to grains of sand) or "a sand." Such checks are helpful in ensuring that the correct functions of words have been assigned.
When words combine to form compounds, the parts of the words that "touch" frequently undergo some sort of change. We can see this in the two English words "synthetic" and "symmetry" which are based on Greek. In the first word, the prefix syn- is combined without change. In the second word, however, the prefix undergoes a change due to the beginning consonant of the second part. If you try to pronounce "synmetry" and consider what your tongue and lips are doing, you may perceive the awkwardness of the unmodified prefix.
Going to phrases, linguists must consider another set of categories. A linguist may have identified the nouns "man" and "rock" along with the verb "to hit," but there is a vast difference between "the man hit the rock" and "the rock hit the man." If a rock and man are involved, we might like to know if it happened in the past or if it represents a future danger. We might also like to know if more than one rock or man is involved. Further, we can express the idea in different ways such as "could the rock hit the man?" and "the man was hit by the rock."
The common concept of a dictionary is a reference book for looking up the meaning of a word. That concept usually restricts the entries and the definitions to the same language. Many people are introduced to >bilingual dictionaries when they study a foreign language. At the same time, they learn the danger of relying on such a dictionary, since a given word "works" only in certain contexts. One story goes that "the spirit is willing, but the flesh is weak" was translated into Russian and back to English as "the vodka is good, but the meat is rotten."
It should come as no surprise that the dictionary has its beginning in the language learning process. Upon encountering a new word, the lexicographer is faced with questions such as: Is this word being used in a new way? What is the difference between these two words that seem to describe similar objects? What other words are typically used with this word? Can I substitute another word for this word? The lexicographer uses the answers to these questions and others like them to build up knowledge that will contribute the to entries in the dictionary.
The first task in writing a dictionary entry is to identify the primary sense of a word, or the first thing that comes to mind when a person hears that word. The primary sense will help to distinguish seconday senses, which are distanced from the primary sense by the lack of certain components. In English, the word "jungle," refers primarilly to an overgrown area of land inhabited by predatory animals. A person referring to another person's yard as a "jungle" is using a secondary sense, meaning that the grass and bushes are overgrown.
A figurative sense of a word is further removed from the primary sense by the limited context in which it may be used. In English, we might say of business that "It's a jungle out there!" but we are referring to neither overgrown foliage or predatory animals. Instead, we are using components of the primary sense of jungle (an untamed, hostile environment) to describe how we perceive the competitive world of business. Although figurative language has considerable expressive power, that power is matched by a potential for misunderstanding, and so it must be used judiciously.
Given a number of senses for a word, the lexicographer must test each of them to ensure that they are neither overly narrow or excessively broad. In English, we might have a number of examples of the verb "to eat," such as "I ate my broccoli," "Expenses are eating us up!," and "He had to eat his words." The first of these is the primary sense: to consume food through the mouth. In the second example, "eat" is used in the sense of consuming resources. In the third example, the sense of "eat" seems to be lost. The meaning is that the person had to retract something that he said. This use of "eat" is an idiom because the meaning is tied to the phrase.
People tend to think of anthropology in terms of the study of "primitive" societies. Since anthropology is generally concerned with human behavior, such studies are relevant wherever people are found. Even though they may not call it anthropology, businesses have been quick to realize the value of cultural awareness as their operations have expanded into multiple countries. The simple fact is that people of different cultures have different ways of making requests, negotiating, trading, giving gifts, showing appreciation, and so on.
Anthropology starts with basic activities. These activities are tied to the common experiences of human beings such as eating, drinking, traveling, working, playing, and sleeping. In categorizing activities, anthropologists move from generic categories to specific ones. For example, one of the prerequisites for eating is obtaining food. A city dweller obtains food by trading at a market; a country dweller raises animals for slaughter; a jungle dweller hunts wild animals. Each understands the idea of obtaining food, but they do not generally consider alternate means.
The study of how people eat is fraught with questions. What do they eat? When do they eat? With whom do they eat? How often do they eat? Do people of different ages eat together? Do males and females eat together? Where do they eat? What is the normal eating posture? What conventions are observed during eating? Who prepares the food? Who serves it? Perhaps, the most pertinent question to start with is: How much does the anthropologist need to know? The answer to that lies in the goals for learning about the culture.
In business, a person will not generally be writing a scholarly paper on the customs of a given country. That same person, however, is probably interested in learning about the culture to accomplish at least two things. First, to gain a favorable reception for their ideas; second, to avoid behaviors that will cause embarassment or rejection. Since business contacts are somewhat limited in scope, they may fairly restrict their attention to proper conduct in meetings, customs at meals, and so on. At the same time, they cannot afford to neglect certain aspects of interpersonal relations which may come into play at any time.
A written work being translated will also have boundaries. Whether the material is literary or technical, it will generally deal with a small subset of the entire range of human activities. Aside from the simple number of activities, the complexity of the translator's task is affected by the distance between the translator's own culture and that of the audience. The task may be further complicated if the material to be translated includes elements that are foreign to both cultures. For the translator, the key to successful anthropological study is familiarity with cultural aspects of the source material and the special treatment they may require in the translation.
A key concept in the work of translation is the genre, or type of work that is being translated. Any given text is an example of at least one genre, although parts of a text may be of different types. For example, fiction usually combines storytelling and conversation. Another example is a speech, which would consist of explaining and motivating elements. A third example is a technical document, which would include procedures and descriptions. Mixing genres indiscriminately can have unintended affects; imagine yourself finding a dialogue between two characters in an appliance instruction boklet.
Genre determines the style of a text. For example, a dialogue in a given language will have rules about how much initial information is required and how the participants are introduced. Similarly, a procedure will assume certain information on the part of the reader while supplying information about the prerequisites for each of the steps and the objects involved. Conversational style will vary by how the participants are quoted and how often they are identified by name. Failure to follow the implicit rules for a genre results in confusion for the reader, even though the words and sentences by themselves may be perfectly understandable.
Whereas technical prose will tend to use straightforward language, other forms of writing will employ a variety of devices. Probably the most common ones are idioms and figures of speech, some of which are so common that people do not think twice about how they might sound to a non-native speaker. Other devices include exaggeration, understatement, irony, and sarcasm. This last of group of devices can create special problems for the translator because their purpose lies in conveying something other than the literal meaning.
Other sources of problems for the translator may lie in the differences of distinction between the two languages. A language may not distinguish between the actions of shouting and whispering, reducing both to the English verb "to say." Another example is that some languages use different pronouns depending on how many people are being addressed, and whether the pronoun "we" includes the speaker or not. In some cases, these distinctions may be merely technical; in other cases, they may directly affect the meaning of the translated text.