Troy Manning

Updated 5/23/09

Determining Word Divisions in an Unwritten Language

When I travel on deputation, I’m often asked about the whole process involved in putting a language into a written form.  The very idea of a language not being in written form is basically completely foreign to us, though we were probably aware that there were some languages in the Amazon and maybe deep in the jungles of Africa that don’t have a written form.  Actually, most of the world’s almost 7,000 languages don’t have a written form. 

But let’s say you are in the deep jungles of Africa, and you wanted to reduce their language to writen form–i.e., you wanted to give them an orthography.  What would you do?  Let’s just talk about one aspect of this process–determining word divisions.  How would you determine what their words look like–i.e., the shape of them on a page?  You definitely wouldn’t just look for the spaces on a page, because the language isn’t on any page. 

In J. Albert Bickford’s book Tools for Analyzing the World’s Languages, he offers guidelines based on phonological word boundaries and syntactic word boundaries.  Phonology, simply put, is the study of speech sounds in the language.  From the perspective of phonology a word is “a minimal utterance.”   Therefore, we can study the phonology of a language to determine word divisions.  Here are his three phonological word boundaries. 

1. No utterance can be shorter than a single phonological word.

The shortest thing you can typically utter in a language is a word.  You wouldn’t spout out a prefix, a suffix, or any other aspect to a language that is shorter than a word–at least, not if you wanted to communicate intelligibly in that language.  (Of course, there are exceptional situations, like at the kindergarten graduation that I attended last Thursday.  The cute, little kindergarteners built the vowel and the consonant houses, and exclaimed the particular sound that each letter “goes.”)

2. Pauses are only possible at word boundaries.

Many of us may get frustrated when a person speaking to us stops in the middle of a sentence and then takes a seemingly long break before finishing, but thankfully, we don’t typically have to worry about their stopping in the middle of a word.  Again, there are exceptions to this, as when you are talking with your mouth full of food and something goes down your throat unexpectedly when you are right in the middle of a word.

3. Look for phonological rules that provide information about word boundaries, then use them to help define word boundaries in unclear cases.

There are phonological rules that apply only within words but not across word boundaries to other words.  Some of these rules are translingual–i.e., they appear in multiple languages.  Maybe that’s because we all have the same instrument to use in articulating our words–a mouth/throat made by God.   In addition, as you study a language, you begin to detect clear examples of phonological rules.   With the interplay of the other principles for finding word divisions, you discover that the phonological rule works only at word boundaries.  Then, when you see this rule working in situations where the other principles might be unclear, this principle may be able to come to the rescue. 

At this point, Bickford moves to syntactic word boundaries.

1. Any phonological word break is generally also a syntactic word break.

This basically builds on our phonological word boundary #1, so it doesn’t contribute a great deal in determining word divisions by itself.  But it is helpful to study how various elements of an utterance can be moved around within a sentence.  As you begin seeing patterns, you begin discovering words. 

2. Any major constituent break (e.g., the beginning or end of a phrase) is also a syntactic word break.

This guideline is helpful in the same way that the previous one was.  When you see larger units in different places in grammatically correct statements, you also discover breaks, and those breaks lead to the discovery of word divisions.

3. Affixes tend to occur next to only  a single type of word (their stems) and in a fixed order; words occur more freely in various combinations with each other.

Bickford is referring to prefixes, suffixes, and infixes.  As you analyze a language, you begin hearing certain sound combinations recurring.  Then you hear patterns in those sound combinations.  These lead to the discovery of pieces of words.  You notice that these pieces of words (e.g., affixes) always occur in a certain order and in fixed positions with only limited combinations.  When you can get your “hands” on entire units of those sound combinations, you realize you have discovered words. 

4. Morphology often shows great irregularities, while combinations of separate syntactic words do not.

This is a complicated principle, but one example of it is with suppletion.  A normal pattern of suffixes may be interrupted by an unusual spelling.  E.g., ‘oxen’ is used instead of ‘oxes.’  But such irregularities rarely occur on the word level.  E.g., the definite article ‘the’ is always going to occur that way.

These guidelines can help in discovering word boundaries, though sometimes they defeat one another since they may lead to opposing conclusions.  In addition, languages have such things as “clitics” which display characteristics of words and of bound morphemes (i.e, units that cannot stand alone as words can).  But at least you can take these guidelines as far as they will allow you to go as you reduce a language to written form.  And maybe some of you will find this quite interesting and challenging and would love to be pointed to the next unwritten language, so that you can give them an orthography which will lead to a translation of Scripture!!  Bibles International is in desperate need of more linguistics consultants.  Translation consultants are hard to find, but they are relatively easy to find compared to linguistics consultants.  Currently, we have only 1, though we are working with 37 different languages!

Updated 6/29/08

The Limits of Dynamic Equivalence in Bible Translation

Notes on Translation No. 121(October, 1987): 1-15

D. A. Carson

D. A. Carson begins his article by applauding the triumph of the theory of dynamic equivalence as a Bible translation philosophy.  He praises its strengths which overcome the “dismal inadequacy of merely formal equivalence in translation.”  He notes the positive impact of the theory of dynamic equivalence in publishing houses and with missiologists and lecturers in Greek and Hebrew. 

Most of his article, however, focuses on the abuses of this theory.  He directs his comments not to “the careful scholars in the field” but instead to the rest of the populace who have pressed the concept “dynamic equivalence” beyond its intended boundaries.  He states his goal as follows: “I am not surreptitiously advocating the overthrow of the principles of dynamic equivalence, but rather encouraging clarity of thought and the adoption of necessary caveats.”

His first caveat is caution about the “limits of equivalence of response.”  Nida emphasizes equivalence of response by the reader as more important than equivalence of form.  Carson recognizes the value of this principle, but he points out the ways in which this principle can be abused, as illustrated by examples from the writings of Charles Kraft.  Kraft advocates “transculturation” which is essentially “the elevation of response over truth.”  Carson notes that Kraft misrepresents those who oppose his ideas, and this misrepresentation is based on a flawed understanding of the dynamics of Jesus’ day.  Jesus’ enemies cannot be equated with those who are fixated upon formal equivalence.  Then, Carson raises the whole issue of exactly what “equivalence of response” really means.  Many misunderstood Scripture when it was first given.  Do we want to reproduce that?  Our definition of “equivalence of response” must be stated in linguistic categories, and our goal must never abandon the truth content of the message.  “The aim of a good translation is to convey the total content-informational, emotional, connotational etc.-of the original message to the reader (or ‘hearer,’ where the translation is read publicly) in the secondary language.”

             His second caveat addresses the limits of equivalence in theologizing.  Basically, he is discouraging the use of the terminology “dynamic equivalence” in areas of ministry that are “far removed not only from linguistic priorities but also from translation itself.”

            His third caveat discusses the “limits to the dichotomy between meaning and message.”  He notes that some have transferred the idea of “dynamic equivalence” to areas outside translation and have then disjointed meaning and message.  He acknowledges the complications that arise with meaning and message between the speaker and the hearer, but he argues that meaning can be very successfully communicated through message.

            His fourth caveat treats the “limits to the equivalence of biblical history.”  He notes that “God has revealed himself to men in time-space history-to particular men and women, spatially and temporally and linguistically located.”  Therefore, we must be very cautious about uprooting revelation from that soil by the way in which we translate. 

            His fifth caveat deals with the “limits to the equivalence of salvation history.”  Not only should we be careful about the historical rootedness of revelation, but we must also be sensitive to the theological connections within Scripture and the progressive-illumination aspect of it.

             In the final section Carson moves to some recommendations.  He disagrees with Nida that “good exegetes and grammarians make poor translators.”  On the contrary, translators don’t usually know enough Greek or Hebrew, nor do they know enough about linguistics, style, or the various aspects of the biblical world.  He ends his article by encouraging the addition of study notes to translations, but only if they are “neutral” and objective.  He also encourages translators never to think that a translation can replace the important function of the preachers and teachers of the Word.  

Updated 3/27/08

English Appreciation 

Many English speakers have difficulties with the English language.  But if we compare English with other languages, we can have much to be thankful for.  First, the English verb is very simple morphologically.  There are basically only two forms for tense: past and non-past.  There is subject agreement marking, but only for singular or plural.  But in other languages, there is also object marking.  In fact, in Kinyarwanda you may need to reflect agreement with three separate objects.  Abaza has agreement markers for person, number, human/non-human, and gender.  Bantu languages mark the verb to agree with the noun class of the subject.  Some also have agreement in definiteness.

Second, the English noun is very simple morphologically.  Noun classification is very simple in English.  There are no genders, except in the case of pronouns.  And there are no cases, except again in the case of pronouns.  Some languages can have about 24 different genders (obviously, they do not correspond to physical gender).  Swahili has 16 different genders, or noun classes.  In some cases, the meaning of the noun helps you determine which class it’s in.  The more reliable indication is the prefix.  However, in some cases, the prefix may be typical of one class, but the noun is actually in a different class.  But the noun triggers the prefixes that are attached to the other elements of the sentence.  So by looking at those other words, you can determine what class the noun is in. 

This morning we looked at Tabassaran, a language spoken in western Russia.  It has 53 different cases.  And we Greek students thought we had it bad with 5!  Many of these cases are actually prepositions that are attached to the noun.  But since they are separate, bound morphemes, they signify different cases. 

So the next time we complain about the phonetics and phonology of English (e.g., how to say the letter combination “ough” in a word), we can at least be thankful that we don’t need to worry much about morphology! 

Updated 3/17/08

Classifying the World’s Languages

In Genesis 11, we have the account of the Tower of Babel.  Though most of the versions say that God “confused” or “confounded” the language of all the earth, I agree with BI’s Chief Translation Consultant that we should probably translate it as “mix”, since God is not the Author of confusion.  And He did quite a good job of mixing man’s language!  It truly is amazing the different ways in which languages express man’s thoughts. 

In Morphology the teacher discussed the main types of languages.  First, there are isolating or analytical languages.  Chinese and Haitian Creole are examples.  These languages have very little morphology.  In other words, syntactical relationships are not expressed by adding morphemes to stems.  Instead, each syntactic function is represented by a separate word.  There are no affixes.  Thus, these languages aren’t productive in the same way English is, where you can simply add affixes to create a new word, like “splendiforously.” 

Second, many languages are agglutinating.  Swahili and Quechua are examples of these.  Several affixes are added to the stem, and each morpheme has its own separate meaning.  Words without affixes are rare.  Words with many affixes are quite common.  In fact, Swahili can have as many as 10-11 different morphemes (e.g., negative prefix, subject agreement prefix, tense marker, root, final vowel, etc.). 

Third, many languages are fusional.  Hebrew, Greek, Spanish, French, and English are examples of these.  The affixes and the bases are fused together.  One morpheme can have several meanings.  The Greek verb form “luo” has the root “lu” plus the suffix “o”, where the suffix indicates person, number, tense, voice, and mood.  I have to write a 10-page paper for this class, and I’ve been studying the Hebrew participle.  It’s interesting to go back and study this language from a linguistics perspective.  Since the participle represents the intersection of the noun and the verb morphology, and since the Hebrew verb system is quite complex, I don’t think I’ll have any problem coming up with 10 pages!

Fourth, many languages are polysynthetic.  Tiwa and Sora are examples of these.  Several stems are combined to form a single word (e.g., a noun and a verb are fused to form one word).  Agglutinating languages have only one stem per word, but polysynthetic may have two that have been fused together. 

More recently, we have been studying what’s called “template morphology.”  Instead of adding one affix upon another, template morphology establishes various slots for the different affixes.  Affixes of different forms can appear in the same slot, but each of the affixes can only appear in a limited number of slots.  Some affixes require that another affix appear with it, though in a different slot.  The template establishes where each affix should go.  The Native American languages evidence this complex system of word formation.

Even if you haven’t followed everything I’ve said, you have probably gotten the impression that the morphology of languages is quite varied and complex.  Languages don’t differ simply in the words used; they also differ at the foundational level of exactly how those words are formed.  And the formation processes between languages can be quite distinct. 

________________________________ 

Eskimos and Snow

Have you ever heard that Eskimo’s have a couple hundred words for snow?  Supposedly, since they deal with snow on a continual basis, they have variously nuanced words to describe the different types of snow.  Well, if you’ve ever passed this story on to others, I would encourage you to let it melt away.  It turns out that it’s just a myth.  Instead, it appears that they have only about a dozen words for snow, not too different than the English language (e.g., blizzard, sleet, slush, avalanche, etc.).  The accurate core of the myth arose in 1911, but it was falsely embellished by a man named Benjamin Lee Whorf, a linguist whose other findings have also been questioned.   Steven Pinker, author of The Language Instinct, dispelled this myth for me, so I’m passing on this helpful news to you.  This is one of the books I have to read for next semester.  I’ve enjoyed the first 74 pages and highly recommend it. 

Last week I finished one of the other books–The Atoms of Language, by Mark C. Butler.  He develops the fact that though languages are quite different externally, they are quite similar internally.  In fact, when languages are boiled down to certain linguistic parameters, it turns out that some differ only by a few parameters.  These slight differences on this level create significant ones on the external level, kinda like just changing a few ingredients in a food recipe (now that I am domesticatized, I can speak authoritatively on this subject!  yeah right!).  Butler’s book is a little more technical, but it would still be interesting for those of you who are more linguistically inclined. 

Only Four Colors

The translator of the Sara Kaba Dem OT in Chad explained to me that they have names for only four colors–black, red, white, and yellow.  All other color names are based on those four or on something in nature.  For examle, blue is “like black” or “like the sky.”  Brown is red-white.  Green is “like the color of the trees.”  How far are they from America’s naming of our colors, thanks to Crayola and the marketing industry (that likes to come up with names of colors that have absolutely no relation to the actual color)!  Imagine trying to translate Scriptures like the one in Leviticus about the leprosy (which is the passage we dealt with during this trip), knowing that you have only four named colors to work with! 

Passivity Not Allowed

During this year’s Translation Consultant Seminar, we were able to work on a translation model that we give to the translators as a reference that they can use as they translate.  The model serves as a “consultant” when the consultant is not able to be physically present.  It tries to represent what the original languages are communicating, but it has to take into account the characteristics of the receptor language.  For one thing, this means that the model cannot contain passive voice verbs.  Most of the languages that we work with do not use this type of verb.  For people who speak such languages, whenever they see a verb, they automatically look for the agent of the action.  What do you do when the agent of the action is unexpressed in a passive voice construction?  This is one of the many challenges that our consultants face as we try to translate God’s Word from one language to another.