Gender in Bible Translation:
Exploring a Connection with Male Representatives1
by Vern Sheridan Poythress
[Published in the Westminster Theological Journal 60/2 (1998) 225-53. Used with permission.]
How do we handle gender in English Bible translation? The articles in World magazine and the subsequent debate in Christianity Today have opened a controversy .2 A special meeting in Colorado Springs on May 27, 1997, convened by Dr. James C. Dobson, produced the “Colorado Springs Guidelines for Translation of Gender-Related Language in Scripture.”3 Those guidelines criticize some of the practices of existing “gender-inclusive” translations. But other people disagree with the guidelines and defend gender-inclusive translation.4 The discussion continues to grow, so that it is difficult to keep track of all its strands.
I propose to focus on one little-noted feature of the debate. A common pattern seems to belong to quite a few of the passages whose translation is disputed. The disputed cases typically involve passages where the Bible sets forth a principle or lesson of general applicability. But it does so using a male human being or a word with a male semantic component5 as the starting point for illustrating and fleshing out the general lesson.
To demonstrate this pattern, we shall have to look separately at several kinds of cases. Then we explore the significance of this common pattern.
By way of preface, let us remind ourselves that translation is difficult. The systemic structures of different human languages do not match one another exactly, either at the level of lexical meanings of words, or at the level of grammatical constructions. Nor do cultures match in the assumptions that they bring to the text. In consequence, translators frequently find that they cannot find any way to translate every nuance of an original text. They try their best, but in the process they make compromises.
Important as these observations are, I cannot now expound on them at length. I must simply assume them. My purpose in this article is to focus on one particular problem, namely male representatives. We will at first focus on certain losses that arise when moving to gender-inclusive translations. Only at the end of the article will we assess whether there are offseting gains.
First, how do we treat generic “he” in English?6 Look at Matthew 16:24-26:
Then Jesus told his disciples, “If any man would come after me, let him deny himself and take up his cross and follow me. For whoever would save his life will lose it, and whoever loses his life for my sake will find it. For what will it profit a man, if he gains the whole world and forfeits his life? Or what shall a man give in return for his life?” (Revised Standard Version, RSV, 1946)
“Any man” in 16:24 corresponds to the underlying Greek word τις, “someone, anyone.” This word may be accurately rendered using the English word “anyone”:7 “If anyone would come after me, he must deny himself and take up his cross and follow me” (New International Version, NIV, 1978). The word “anyone” shows that both men and women are included. Translational adjustments of this kind are not in dispute. But the verse still contains generic “he,” in the form of “he” and “his,” referring back to “anyone.” Some people find this usage distasteful. So the New International Version Inclusive Language Edition (NIVI, 1996) undertakes to eliminate it:
16:24 Those who would come after me must deny themselves and take up their cross and follow me. 25 For those who want to save their lives will lose them, but those who lose their lives for me will find them. 26 What good will it be for you to gain the whole world, yet forfeit your soul? Or what can you give in exchange for your soul? (NIVI)
In verses 24-25 all the singulars, “he/his/ him/himself,” are converted to plurals in order to eliminate generic “he.” In verse 26 the NIVI adopts a different strategy. It replaces the third person singular “he/his” with the second person “you.” The New Revised Standard Version (NRSV, 1989) uses “they/them/their” in all three verses, while the New Living Translation (NLT, 1996) uses “you” in all three verses.
Changes like these are not exceptional. Because generic singular is a convenient and frequent usage in the Bible, the NRSV, NIVI, and NLT end up using “they” and “you” in a large number of passages where earlier translations had generic singular “he/his/him.” In still other instances, the new translations adopt passive rather than active constructions, or substitute descriptive nouns for pronouns in order to avoid using “he.”
Now, let us be clear. The gender-inclusive translations still achieve a rough approximation of the meaning of the original when they change the pronouns. But it is an approximation. When we look at finer nuances, shifts from singular to plural and from third person to first or second person result in subtle alterations. “You” speaks to the hearer directly,8 “we” speaks to the hearer and includes the speaker, while “anyone/he” pictures a general case “out there,” as it were. The relation between a statement and the addressees subtly changes when we shift from third to second person. The relation to the speaker changes if we shift to first person.
“He/his/him” refers to a representative individual person within a group of people, while “they/their/them” refers to the plurality of members of a group. In changing from one to the other the focus shifts from the individual to a plurality. It is a change. Of course, the exact character of the alteration may differ from passage to passage. But the alteration always produces some difference in meaning.
It is often claimed that no harm is done, since the original meaning of the text is still implied, directly or indirectly, in the translation. A statement about a plurality using “they” still implies an application to each individual (“he”) within the collectivity. Conversely, a statement about a single sample member using “he” implies truths concerning the plurality of all members of the collectivity. Similarly, readers can infer a general truth from something that explicitly addresses “you.”
But in reality the people who argue this way have already conceded that the meaning has subtly changed in their translation. An explicit semantic content in the original has to be inferred in the translation, while what was only inferable from the semantics in the original becomes explicit in the translation. The shift from direct statement to inference is significant. It is a subtle change in meaning. Moreover, a shift from “he” to “you” results in a subtle shift in focus, even though we may express the same general truth through both modes of expression. When we use “he,” the pictorial starting point is a sample person who is mentally positioned “out there.” When we use “you,” the pictorial starting point is inviting the addressee to picture himself in the situation. In English, we possess something like a generic “you.” I might say, “If you get in trouble with the law, you will suffer for it.” In the larger context, my addressee may be in no imminent danger of falling afoul of the law. I might not be worried about him. I am making a general observation about anyone. But by using “you” I invite the addressee more directly to place himself in his imagination in the shoes of the potential lawbreaker.9 It is easy to exaggerate the differences between these several modes of expression. But differences there are—differences in nuance in the total meaning-impact, not merely differences in phrasing with no meaning difference. After extensive theoretical reflection on meaning and translation, Ernst-August Gutt comments, “Translators should have a firm grasp of hitherto neglected aspects of meaning. In particular, they should understand that there are important differences between expressing and implicating information, between strong and weak communication.”10
To appreciate fully what Gutt says, biblical scholars have to shift their point of view somewhat. Many biblical scholars spend most of their time thinking and writing about the theological value and interpretive implications of the passages that they study. They write commentaries whose main business is to make explicit the many implications of the text. If two wordings leave the theological implications the same, they are equivalent from the scholar’s point of view. But literary stylists and linguists studying discourse focus on other aspects of the text. They would note that subtle differences exist between explicit and implicated information, direct and indirect address, active and passive constructions, second person and third person discourse. These produce subtle nuances in the meaning-texture of the total act of communication. Translation into another language never succeeds in conveying absolutely all of such nuances. But the faithful translator endeavors to do so as far as possible.
Translators console themselves by saying that “all translation is interpretation.” They are right. The most accurate translation can only be accomplished when we thoroughly understand the meaning of the original, including all its nuances in all their dimensions. Translators must first interpret the passage thoroughly in the original. Only then are they ready to produce a translation that conveys not only the main meaning but all the nuances of the original.
But the motto, “all translation is interpretation,” is turned into another meaning if we then use it as a blanket justification for rewriting the text in the way that an interpretive commentary would do. An interpretive commentary expounds the implications of a text, and makes explicit what the text leaves implicit. Such has not generally been the job of mainstream translation. But the American religious public has become lazy about the Bible and busy with other affairs. Today ordinary people do not read commentaries. They seldom read the notes in a study Bible or the footnotes in modern translations that alert readers to translation problems. So we may try to include the extra information in the text explicitly, in order to make it easy for them. We paraphrase. We explain metaphors in ordinary prose. We expand tightly packed theological exposition. By doing so, we help readers to understand the Bible much better. Many benefits accrue, similar to the benefits of commentaries and the benefits from notes in the margins of study Bibles. But if we produce interpretive paraphrases, and then label them “The Bible” and “translation,” we have blurred the line between translation and commentary in an unfortunate way.
Whatever we think of the result, we must remain clear about the meaning of the original. That meaning includes not only “basic content,” but nuances arising from style, focus, emphasis, allusion, metaphorical color, literary form, thematic structure, rhythm, tone, register, literary density of information, directness and indirectness, explicitness and implication, and intertextual connections, to mention a few complex dimensions of the whole. In translating generic singulars, something is lost or changed through pluralization, shifts in person, and passivization.11 The debate can then arise over whether some gain compensates for these losses.
In addition the changes introduce possibilities for distortion and misunderstanding. For example, in Matthew 14:24 (NIVI) “their cross” in the singular could be construed as a single cross belonging to the whole group of followers jointly. The group jointly has responsibility for a single “cross,” a group shame. They also have a group life, in which they deny “themselves,” their former identity as a group. The focus subtly shifts from individual to group. Likewise, if nothing but the second person “you” occurs, it may become less clear whether the saying applies to all human beings, or just to the immediate addressees.
We may further illustrate the difficulties with other passages. Compare two versions of John 14:23:
If anyone loves me, he will obey my teaching. My Father will love him, and we will come to him and make our home with him. (NIV)
Those who love me will keep my word, and my Father will love them, and we will come to them and make our home with them. (NRSV)
The NRSV substitutes plurals for the generic singulars found in Greek and in the NIV. But this step results in an unintended ambiguity in the product. The last clause, “make our home with them,” has a plurality of people, “them,” combined with a single dwelling place, “our home.” Conceivably, it might mean that the Father and the Son make a home with each person. But it might also mean that the Father and the Son make a single home with the plurality of people together. That is, they come and dwell with the church corporately. It may be that this latter interpretation is closer to the surface or more “obvious” than the first, since it responds to the difference between the singular “our home” and the plural “them.” Such a thought of corporate dwelling is genuinely biblical, as 1 Corinthians 3:10-15 and Ephesians 2:22 show. But it is not the thought found in John 14:23 (Greek text). Both the Greek and the NIV picture the Father and the Son making a dwelling with each person, not with the church corporately.
Similarly, the clause “my Father will love them” may be understood either corporately or individually. Does the Father love “them” by loving each individual, in his individuality? Or does the Father love “them” by loving them as a group, loving the church? Both readings are theologically true, as Ephesians 5:25 and Galatians 2:20 indicate. We could tentatively infer either truth from the other. But, unless we use further information, neither inference is really solid.
Love is a multifaceted personal action, with many dimensions and many possible variations in nuance. The connotations change when we shift from loving “him” to loving “them.” Consider an example from human experience. Mary may love every person belonging to a certain social group, yet not love the group because she disagrees with its purposes. Conversely, she might love the group, yet not love some of the people in it. Mary is glad that John belongs to the Socialist Party and furthers its cause, but she has no taste for intimacy with John outside the group. She does not want to be his confidante.
“Mary loves those who work for her favorite charity, the Philadelphia Literacy Institute.” But the list of workers is long, and Mary does not even know all their names! “Mary loves each person who works for the Philadelphia Literary Institute.” She goes to chat with each of them every day. The two modes of expression suggest two loves, subtly different from one another. The two modes also tend to suggest two subtly different reasons for the love. In the first case, the love is based on attachment to the charitable institute. In the second case, the love builds on acquaintance, and may have many motives unrelated to the institute. Of course I heighten the difference between the two kinds when I add further commentary about Mary not knowing their names or chatting with each one. But I believe that a subtle difference in atmosphere is already present even without the commentary. My point is a subtle one, and so it is hard to make without exaggerating it.
Consider now the translation of Revelation 3:20.
Behold, I stand at the door and knock; if any one hears my voice and opens the door, I will come in to him and eat with him, and he with me. (RSV)
Listen! I am standing at the door, knocking; if you hear my voice and open the door, I will come in to you and eat with you, and you with me. (NRSV)
The NRSV has changed “he/him” to “you.” The NRSV again presents us with an ambiguity. The NRSV may mean that Christ will eat with each individual “you” who hears his voice. But it may also be a call for the whole church corporately to hear his voice, corporately to repent. In response Christ then promises to eat with them corporately. The “you’s” in the preceding context of Revelation 3:14-19 all address the angel, and through the angel the church at Laodicea. The shift to the individualizing “anyone” is lost in the NRSV. And the loss may be significant. In a good many of the exhortations to repentance in Revelation 2-3, Christ calls individuals to repentance, even in the face of the possibility that others in the church do not repent (note, for example, Jezebel in Rev. 2:21, and the distinction between the “few” and the many at Sardis, Rev. 3:4; note also “he who overcomes …”). The implication of Revelation 3:20 is that each one is to hear his voice, even if others in the Laodicean church fail to hear. The NRSV fails to convey this significant dimension of meaning.
I have heard people observe in response that American Christians are far too individualizing. The danger, they say, is not of missing an individualizing note, but of missing the corporate dimension of NT Christianity. I would agree with this assessment of American Christianity as a whole. And I would agree that this fact decreases the practical damage done when verses are altered. But I would not agree with another possible implication, namely that we can safely reduce the individualizing aspect of these particular texts. In the context of doing translation, this sort of argument is an embarrassment. Readers’ problems with other texts and with other teachings of the Bible must not become an excuse for a loose attitude toward translating these texts. Moreover, not all English-speaking readers are the same. The translator does not have the luxury of addressing different groups according to their different problems. Even if he did, it would be paternalistic for him to decide what he thinks is “good for them” and then alter nuances of the biblical text accordingly. He would thereby deny them to some extent the freedom to make up their own minds about whether NT Christianity is really an individual or collective institution, and in what ways it is so. Such a translator presumes too much.
Then why have the changes been introduced? In 1998 generic “he” no longer occurs as frequently as it once did. Writers have tried to find work-arounds. But generic “he” still occurs in the secular press.12 Extremely skilled and sensitive writers in our day still use generic “he” on occasion because they find that that is sometimes what they want to say, and they would lose nuances by adjusting to plurals or first or second person.13 The NRSV, NIVI, and NLT all use generic “he” in Psalm 109.14 In the fourth edition (1990) of the book On Writing Well, William Zinsser eliminated many of the generic masculines that occurred in earlier editions of his book, but he also says, “Where the male pronoun remains, I feel that it’s the only clean solution.”15 Of third person plurals Zinsser says, “I don’t like plurals; they weaken writing because they are less specific than the singular, less easy to visualize.”16 “A style that converts every ‘he’ into a ‘they’ will quickly turn to mush.”17
Now, in the context of Bible translation one ought not to tolerate these losses as long as a way exists of avoiding them, namely by using generic “he.” Modern writers have authority over their own meanings, and can alter them if they choose. They can rephrase or restructure what they are saying in order to eliminate all generic masculines. They can convert their styles to “mush” if they like. But the translator does not have the same authority to introduce subtle alterations in the meaning of the biblical text.
People have sometimes pointed out that the New Testament writers and the LXX as well sometimes make shifts of person or number in dealing with the Old Testament. Does not this fact constitute an endorsement of the practices in the NRSV, NIVI, and NLT? I do not think so.
The NT writers in their quotations are more like preachers making an application than like translators. I myself would feel free to alter a quotation for the sake of application, if I knew that my hearers had a translation available. I would not be claiming to give the most accurate translation for general purposes, but rather an interpretive rendering that brought out some of the implications of the original. The NT offers interpretive renderings rather than a uniform model that endorses a particular brand of translation. What about the LXX? It is not flawless as a translation. It served its purpose reasonably well, as do most English translations with their flaws. But when the flaws are systematic, as they are in the case of the existing gender-inclusive translations, further analysis is necessary.
Now what is the significance of the use of generic “he”? “He” is used generically, that is, to speak of a sample individual to whom a general principle applies. The general principle typically applies to both men and women. In this sense, “he” encompasses both men and women; it is inclusive. But is it truly “gender neutral”? That is, does there remain no connotation of “male” deriving from the masculine gender? Reality is more complicated.The American Heritage Dictionary perceptively comments:
If he were truly a gender-neutral form, we would expect that it could be used to refer to the members of any group containing both men and women. But in fact the English masculine form is an odd choice when it refers to a female member of such a group. There is something plainly disconcerting about sentences such as Each of the stars of It Happened One Night [i.e., Clark Gable and Claudette Colbert] won an Academy Award for his performance. In this case, the use of his forces the reader to envision a single male who stands as the representative member of the group, a picture that is at odds with the image that comes to mind when we picture the stars of It Happened One Night. Thus he is not really a gender-neutral pronoun, rather, it refers to a male who is to be taken as the representative member of the group referred to by its antecedent. The traditional usage, then, is not simply a grammatical convention; it also suggests a particular pattern of thought.18
Does the same thing happen in Hebrew and Greek? Hebrew and Greek have gender systems of their own. The two languages differ from one another as well as from English. In Hebrew and Greek nearly all nouns and adjectives, not simply third-person personal pronouns, are marked for gender. One consequence is that in Hebrew and Greek gender as such is not so tightly linked to male and female. In Greek, ἀλήθεια (“truth”) is feminine and θάνατος (“death”) is masculine. But native speakers did not therefore conclude that truth was literally female or death male. It is a convention.19 Of course, when pronouns20 refer to human beings, the gender usually lines up with sex. Masculine is the “default” gender in referring to groups of human beings of mixed sex or of unknown sex. The listener’s instincts are to try tentatively to identify the sex of human referents by the gender marking of personal pronouns. Hence, the listener to some extent “pictures” a male figure on the basis of a masculine personal pronoun. The context decides whether the reference is literally to a particular male human being, to a group of males, to a male as an illustration or sample from a group, and so on. These constructions appeared within the context of patriarchal cultures, where it was natural to take a male person as a sample, as representative of anyone within a larger group. The usage “suggests a particular pattern of thought.” It is easy to exaggerate the significance of this pattern. Gender is not identical with sex, in Hebrew or Greek or English. But when we are dealing with pronouns referring to human beings, the correlation between gender and sex is always there.
Thus we may state as a general observation that, in eliminating generic “he,” gender-inclusive translations endeavor to retain the general principle expressed in a verse. But they do not express the fact that the principle is expressed using a male sample, a male representative who embodies or illustrates the principle in operation.
“Man” for the human race
Consider now the use of “man” to designate the human race.21 “Man” designates the human race in contexts like the following quotations from the secular press: “Early Man’s Journey out of Africa” (1995); “where it [Homo erectus] may have evolved into modern man” (1995); “Clean air and ozone obey no manmade boundaries” (1997).22 While such a use still exist in English, there are distinct advantages to using it in Bible translation. Though it is disputed, Genesis 1-5 in its use of words and names links the naming of man, woman, and the race with the headship of Adam. The matter comes into some particular focus in Genesis 2:23 and 5:2, because of the explicit mention of naming. The result is that these passages in their foundational character create the potential for subtle connotative resonances with the total thinking of Hebrew speakers about sexuality. They create further potential resonances with later usage in the rest of the OT canon, as people consider the later canon in the light of the earlier. The total complex generated by this interaction is subtle, complex, and elusive.
Modern scholars can attain only an approximate understanding of the full complexity. But even an approximate understanding shows that we cannot capture everything in translation. The word “man” in English, used to designate the human race, is not an exact equivalent to adam in Genesis 1:26 and 5:2. But I cannot find anything better. Moreover, I think it is easy to overestimate the alleged problems with using “man” with this meaning. For the most part, it is not that people do not understand such a meaning, but that they do not like it. Neither would they like what Genesis 1-2 and 5:2 do in the original Hebrew—for more or less the same reasons. In both Hebrew and English a term with male connotations designates the whole human race. This usage resonates literarily with the context in Genesis 5:2 and the context in Genesis 3, in which Adam is representative of humanity. His obedience or disobedience has consequences for all his descendants, as Paul states in Romans 5:12-21. It is therefore fitting that his name should match the name for the human race in Genesis 1:26-27 and 5:2. Gender-inclusive translations, while preserving the main point, leave out the connotation of a male representative by translating Genesis 1:26-27with “humankind,” “human beings,” or “people” instead of “man.”23
Next, consider the translation of ’ish ( אִישׁ ).24 ’ish does not always mean “man.” ’ish occurs in idiomatic constructions with the sense “each one” (e.g., Lev. 15:2; 22:4; 22:18; 1 Chron. 16:3; etc.).25 The main problem is that gender-inclusive translations on the market eliminate male marking in other passages where they have no lexicographical warrant.
Some of the cases are nuanced and invite further discussion. Consider Psalm 1:1, “Blessed is the man who does not walk in the counsel of the wicked or stand in the way of sinners or sit in the seat of mockers” (NIV). NRSV, NIVI, and NLT change “the man” to “those.” The English now runs, “Blessed are those who …,” or a similar phrasing. The change from singular to plural presents us with the same kinds of changes in nuance that we have already noted. The plural is, in Zinsser’s words, “less specific …, less easy to visualize.”26 Moreover, with the singular, the reader tends to picture a single man standing against a multitude of wicked people, sinners, and mockers. The contrast between the single man and the plurality of sinners, present in Hebrew, simply drops out when we convert the singulars to plurals.
In addition, Psalm 1:1 starts with the picture of a person who happens to be male. The native speaker of Hebrew reads הָאִישׁ, “the man.” Nothing in the immediate context overturns the instinct to assign tentatively the meaning “the man,” and to think first of all of a male human being rather than a female. The native speaker knows, in the back of his mind, that a masculine rather than a feminine term is likely to be used in a context where the author wants to talk about a sample human being from within a group composed of both sexes. The sex of the sample person may or may not be germane to the point that the author wishes to make. That is, the reader must determine from the larger context whether the sex of the sample person functions to qualify the range of application of the sentence (in Deut. 22:13, Ps. 80:17, and Prov. 6:27 it does; in Ps. 1:1 it does not). The native speaker therefore holds open the range of application. But meanwhile, partly because of the possibility that the text may actually be specializing to male human beings, the reader pictures a man, a male. This male is then a representative for a truth applying to a larger group. But temporarily the exact composition of the group remains undetermined. After reading Psalm 1, sensitive readers know that the Psalm as a whole offers the “man” as a representative, an ideal, to be emulated by readers. He is a model for both men and women, and for children as well. Implicated meaning includes application to many. But the starting point is the picture of one, and that one is male. The maleness is not essential to the main point that the Psalm makes. But it is there, as a semantic component as well as grammatical gender, to the original readers. The gender-inclusive translations simply eliminate this semantic component.
One can make similar observations about many of the case laws in the OT. Quite a few case laws begin with phraseology such as “if a man,” using the word ’ish (for example, Exodus 21:13, 14). Within the cultural context of the Old Testament, the case laws are for the most part formulated in terms of a man’s rather than a woman’s point of view.27 In such a context, an introductory ’ish posits a representative case. “A man” represents the kinds of things that may happen to anyone. The principle includes women. But the explicit formulation has a male starting point. The text mentions a man explicitly, in order that we may understand and infer from this particular case general principles of equity. The male marking thus belongs as a semantic component of the original. If we preserve the meaning distinction between explicit and implicated information, we try to convey the marking in English.
Now, NRSV, NIVI, and NLT regularly lose this male marking. They contain a formulation that expresses the general principle of equity, and that, surely, is the main point. But they drop one aspect of the meaning, by not expressing the subtle interplay between a male representative on the one hand, and a general principle applying to both men and women on the other. They change the interplay between explicit and implicated information. The alterations do change meaning. A subtle change, but yet a change.
Sometimes the fussiness involved in suppressing the male marking is faintly ludicrous. Exodus 21:18 describes a situation in which men quarrel, hit one another with a stone or fist, and end up with serious injury. Though it is physically possible for women to quarrel thus, the physical violence in the picture reinforces the tendency for ancient readers to picture, as a representative instance, two men. But in NRSV, NIVI, and NLT it is two “individuals” or “people.” In Exodus 21:22 it is no longer two men fighting, but two people. In Exodus 22:5 it is no longer a “man” who lets his livestock loose, but “someone”—even though in OT times men would be far likelier than women to own livestock and to be out in the fields managing them.
What do we do with anēr (ἀνήρ, “man”)? When should we translate with the English word “man”? A thorough analysis of meaning involves great complexities. In part, the problem is the classic problem of lexicographical analysis. In learning a language, we infer the meaning of particular lexemes from the context of their occurrences in sentences, conversation, and social settings. But not every meaning of the context can we ascribe to the particular lexeme that occurs within it. How do we separate between the contribution of the context and the contribution of the lexeme? It is not easy. In fact, ultimately an absolute and strict separation is impossible.
The problem is easily illustrated by the occurrences of anēr in the NT. In some contexts, of course, anēr means “husband.” But if we eliminate these, most of the other contexts do not conclusively show that anēr has a semantic component of “male”; neither do they show that the word lacks the semantic component of “male.” The average use where anēr refers to a male human being will not serve as conclusive evidence, because a word meaning “person” could also be used in such contexts. The maleness of the person in question would then be inferred from the information in the context, but would not be contained in the word itself. Equally, the examples that some people cite in the other direction are not conclusive, because there are other possible readings of the situation.
For example, Matthew 14:35 depicts a situation “when the men (anēr) of that place recognized Jesus.” Some have claimed that both men and women were involved. But how do they know it? Perhaps the men as opposed to women were exclusively or more prominently involved in the process, so that Matthew decided to focus on them.28 Ephesians 4:13 uses anēr is a strikingly creative (and metaphoric) way, probably in relation to the thought that Christ is the perfect man. So it is not clear that the male marking has disappeared from the base meaning on which the metaphor builds. Are the men of Nineveh singled out in Matthew 12:41 because within the ancient world men would have the connotations of the more weighty witness in court? Or because the men are representative of all? Or is there a special usage here, with plural andres plus a place limitation, that would cover men and women? This last possibility does deserve more study. And so it goes with some other passages.
The fact is that in this case, as in many other cases of lexicographical analysis, there is considerable scope for reading evidence in more than one way. Now, this situation would not be so bad were it not for the fact that people on all sides of the modern cultural controversies have a heavy emotional stake in issues of sexuality. It is therefore painfully easy for people to read in their own biases. I nevertheless believe that the burden of proof is on those who claim that in a particular context anēr has lost all of its male semantic component. Why? Because this semantic component is definitely there in some of the occurrences, and the existence of anthropos as a more neutral designation fills the need to talk about mixed groups, individuals whose sex is unknown, and so on. It is linguistically improbable that we would find anēr moving toward near synonymy with anthropos in many contexts, leaving Greek with no obvious, convenient term to use when one wants to specify that one is talking about male human beings.
One other principle of semantics needs consideration. Many lexemes have more than one sense. Even for those lexemes where one sense in dominant, there may be special idiomatic constructions that bear an altered or extended sense. And there may be contexts where some normal semantic component of the sense is neutralized. For example, though the Hebrew בֵּן in the singular usually bears the sense “son,” it can be used in the plural to designate a group of children of both sexes, and it can be used in idiomatic constructions like בֶּן חַיִל, “mighty man.”29 But we expect that nearly always contextual clues will supply indications as to which sense or which idiom is being used at a particular spot. (Thus, the plural form of banim is a contextual factor indicating the possibility of absence of a male semantic component.) Differences in sense are thus controlled by context. Ambiguities in meaning are usually resolvable through context, though any language supplies occasional cases where a speaker wittingly or unwittingly leaves an ambiguity that cannot be resolved from context, and fails to make himself clear.
Now, what happens when we apply these insights to anēr? Some people would claim that anēr sometimes means “person” or “people” without a male semantic component. But most people would still admit that elsewhere (and not just where it means “husband”) anēr sometimes carries a male semantic component.30 That is, according to these people, sometimes anēr means (roughly) “person,” sometimes “male person, man.” Can we specify contextual clues that determine which of these two senses occurs in any one particular place? If not, we are making the very implausible claim that the two different senses occur in free variation. But then the next generation learning Greek would never learn that anēr has a male semantic component at all, since the contexts never distinguish this hypothetical sense from the more general sense “person.”
Now make it a little more complex. Suppose we were to say that the only contextual clue to a male semantic component is a context where in fact male human beings are being referred to. It is then the context, either textual or situational, that contains the information about maleness. Again, how would language learners be able to distinguish this situation from one in which anēr simply meant “person” and the context indicated maleness?
I conclude that it is far more likely that the “default” sense for anēr includes maleness. If some occurrences of anēr do not include this semantic component, it is some special factor in the context that neutralizes it. The lexicographer then tries to describe just what contexts lead to neutralization. Such contexts, if they exist, will probably be specialized contexts.
But if we look at what the existing gender-inclusive translations have actually done with anēr, there seems sometimes to be a far looser treatment. In Acts 1:21 Peter discusses replacing Judas: “Therefore it is necessary to choose one of the men (anēr) …” (NIV). But in the NIVI and NLT “men” disappears to become “one of those” (NIVI) or “someone else” (NLT). This text is theologically significant because it indicates that Peter did not think that a woman could be an apostle.
In Acts 20:30 Paul warns the elders at Ephesus about false teachers: “Even from your own number men (anēr) will arise and distort the truth …” (NIV). Indirectly Paul indicates that the elders were all men. This theologically significant detail drops out in the NRSV, NIVI, and NLT. We hear simply of “some” who will arise (NIVI), or a similar expression. Note also that NLT eliminates any male element in translating Matthew 7:24,26, Luke 14:24 (compare 14:20), Luke 22:63, Acts 8:2, 11:20.
The fact is that within modern culture, with its highly egalitarian tendencies, we no longer understand very easily all the reasons why a male markedness might be used within the ancient world. We therefore have a tendency to underestimate the probability of male markedness in particular occurrences of anēr. This underestimation is exhibited, I believe, by the existing gender-inclusive translations.
How do we translate ὑιός (huios, “son”)?31 Some of the NT usages have, I believe, been influenced by the Hebrew use of בֵּן/בָּנִים. But in general huios has a male semantic marking that τέκνον (teknon, “child”) lacks, even in those passages like Romans 8 where both terms occur.
The problem here is somewhat similar to the problem with anēr. And my observations are also similar. Given the existence of teknon, the burden of proof is on someone who claims that huios has lost the semantic mark of maleness—apart, perhaps, from some special idioms. Consider: is it likely in English that in the future “son” would come to be virtually synonymous with “child,” and we would be left with no convenient term for designating a child that we want to specify as male?
What about the use huios in Galatians 4:4-7? The context of Galatians 3:26-29 shows that in Christ both men and women can become sons of God and sons of Abraham (note also Gal. 3:7). Generations of Bible interpreters have always understood Galatians 4:4-7 as including both men and women within its scope. But Galatians 4:4-7 uses “son” in the context of a live metaphor. In Galatians 4:1-2 the picture of a son inheriting an estate becomes the model for understanding that believers inherit God’s spiritual blessings (Galatians 4:3-7). Just as a literal biological son has a time of minority and then enters into his inheritance, so also, through Christ the unique Son (Gal. 4:4, 6), we have had a time of minority and then enter into the status of sons, metaphorically speaking. A male marking still belongs to the lexical item “son”; the context shows that the use is metaphorical and that it applies to both men and women.
The meaning is subtly changed if we merely translate with “child/children,” eliminating the male marking in the original. The result is a similar meaning, in that now men and women are declared to be children of God through Christ the Son. But this similar meaning is not identical to the original. A nuance have been changed, from an original meaning where we, men and women alike, were declared to be sons. The main theological conclusion to which the metaphor points is the same whether we use “son” or “child.” But the basic picture on which the metaphor builds is somewhat different in the two cases, because in the second translation (“child”) we have eliminated a semantic component that is present in the original.
Some people may still say, “But what practical difference does it make?” We should translate the word of God in a manner that is faithful in every respect, not just in a minimal way. The minimum is not our goal. We ought not to be satisfied with a rough translation, of which we may only say that we think it is close enough for our most immediate practical purposes. We aim at maximal representation of every aspect of meaning.
Let’s look at another aspect of this problem. Suppose, for the sake of argument, that it truly “makes no difference” which way we translate. Why not translate it using “son”? Ah, there’s the rub. Such a translation is now thought to be “offensive” or “insensitive.” It might be misunderstood as excluding women. Is it really likely to be misunderstood, when read after Galatians 3:26-29? Can’t even a fairly dull person see that we have a metaphor? But this concern for misunderstanding still remains. We shall return to it below.
Romans 8 contains another example. The words teknon (“child”) and huios (“son”) occur scattered throughout the passage, with no apparent reason why “child” occurs in one place and “son” in another. Surely the main theological points could be made using either word. Both words are used metaphorically, to speak not of a social and legal relation to a human parent, but a Christian’s relation to God.
Since Romans 8 oscillates between teknon and huios, some people have argued that in this context the two terms are synonymous and should both be translated “child(ren).” But this argument is fallacious. Let us consider a related example. In a single context I may talk about God as both master and king. “My Master commands it; my King bids me do it.” What is the difference if I say, “My King commands it; my Master bids me do it”? There is very little difference in the overall point. But my usage does not show that “king” and “master” are synonymous. I use both terms metaphorically. In both cases I refer to the same overall reality, namely God’s authority and right to command, and my submissive, obedient relation to him. But the starting point for the two metaphors is subtly different. In the one case the starting point is the picture of an earthly king; in the other, of a master in relation to a servant or slave. We do not preserve this distinction if, in a translation into another language, we used a word meaning “king” to translate both statements.32 Similarly, I can say, “I am a slave of Christ, a servant of Jesus my Lord.” The virtual interchangeability of “slave” and “servant” does not imply that they are synonymous, either in English in general or in this sentence in particular. “Slave” in its base meaning indicates a relation more comprehensive and less easy to free oneself from than the word “servant.” The differences in meaning remain even in a context where both terms are used in a metaphorical sense. Or again, “My employer acts like a general toward his employees. He orders them about like a ship’s captain.” “General” and “ship’s captain” are not synonymous. The one is a ruler in an army, the other in the navy. I use both terms in a simile, but the simile does not destroy the difference in meaning in the two comparisons.
From a linguistic point of view, oscillation between two terms proves nothing except that, in a particular context, an author can oscillate. Several factors may be behind this oscillation. First, the two terms may indeed be synonymous in meaning in nearly every respect. Second, the two terms may be used to make two distinct but related points. Third, the two terms may have differences in lexical meaning, but be used in a context where the differences are neutralized. Fourth, the two terms may have differences that are still present in context, but that make little difference in the overall thrust of what is being said. The fourth possibility is the one illustrated in the examples above. In such cases, translation ideally preserves the nuances involved in the distinct senses, and preserves distinct metaphors and similes arising from the meaning differences.
In Romans 8, huios continues to possess the semantic component “male” in the base meaning on which the metaphor builds. So there is good reason to translate using “son,” and to preserve in English the shifts between “son” and “child” that mirror what we find in Greek. Moreover, there is a further bonus to preserving the word “son.” The model of Christ the Son, visible in Romans 8:29, resonates with the occurrence of huios (“son”) elsewhere in Romans 8. Retaining the translation “son” as the English rendering of huios enables us to retain the connection between Christ’s Sonship and our sonship. The connection recedes if we consistently translate using “child.”
What is happening in Romans 8 and in Galatians 4:4-7 is similar to what we have already observed in other cases. A term with male marking is used to represent a group including both men and women. The usage in Romans 8 and Galatians 3:26-4:7 appears to rest partly on the fact that a single male, namely Christ himself, has represented us and achieved deliverance for us.
What about the usages of huios influenced by Hebrew, such as (possibly) Matthew 5:9 and 9:15? It is still valid to distinguish between Hebrew on the one hand and Greek “influenced” by Hebrew on the other. The influence, not only from the LXX specifically, but from bilingualism generally, makes it possible to use huios in an extended, expanded, or semimetaphorical sense matching the Hebrew ben/banim. When someone does that, he is quickly understood by others familiar with the Hebraic influence. But such an extended use does not cancel out what the native speaker knows, namely that huios when used less “playfully” means “son,” not “child.” I think, then, that even in the instances of extended use there is a background in which the native speaker knows that male marking is part of the base meaning.
We preserve this complex interplay between a base meaning and an extended sense when we regularly use “son” as the English translation. In the cases where the English NT has an extended usage, everyone knows that it is an extended usage, and easily adapts, all the while knowing that the base meaning of “son” continues. The relation between extended use and base meaning in English mirrors the analogous relation in Greek. All this worked in English until the feminists invaded.
We get a similar phenomenon in pre-1960 English, where “brother” was used metaphorically for Christians. Christian congregations would be addressed as “brethren,” as they are still addressed as “beloved” or “loved ones.” Among some Christians, this usage was so well established that most of the time no one thought about it. But if asked, everyone would have told you that, of course, this was a special use, a metaphorical use, and that “brother” basically meant “male sibling.” The extended use, in direct imitation of English translations, which in turn imitated the Greek NT, did not result in any broadening of the base meaning to eliminate its male component.
A quick overview of Greek and Hebrew vocabulary shows the fundamental difference. Greek has a three-way contrast between ὑιός (“son”), θυγάτηρ (“daughter”), and τέκνον (“child”). Hebrew has a two-way contrast between בֵּן (“son”) and בַּת (“daughter”). Hebrew does have the word יֶלֶד, “child.” But as far as I can see, it normally applies to youngsters only, not to grown-up sons and daughters. τέκνον applies to adults as well. In Hebrew, בָּנִים must cover mixed groups that include adults, because there is no convenient third term. In Greek, the term τέκνον does the job, so there is no pressure to make ὑιός serve this purpose and produce contexts where the male semantic component is neutralized.
Translating words for ancestors
What should we do to translate πατήρ (“father”) and אָב ?33 In many uncontroversial contexts, “father” is a good equivalent in English, used by virtually all the translations. But some contexts are more difficult. For example, Hebrews 11:23 refers to Moses’ parents as πατέρων (plural). We need the translation “parents.” Moreover, both the Greek and the Hebrew terms can refer to more distant ancestors as well as first-generation ancestors. So far so good. But the tendency in existing gender-inclusive translations has been to obscure the patriarchal character of OT thinking about family lines. In many of the salient usages, ’ab is not simply “ancestor,” but something more like “forefather.” In a subtle way some focus or greater prominence belongs to male ancestors. Through these males the tribal and clan structures are traced. Gender-inclusive translations have sometimes eliminated this male component.
The larger context of language and culture
So far we have considered particular translation problems. But we cannot ignore the larger context. And then we confront a host of complex questions. What is the current state of English-speaking cultures and subcultures? Is this state good or bad, and in what respects? What are the effects of modern feminism, egalitarianism, consumerism, and uncritical love of “progress”? How are people now reading the Bible? How much, and in what contexts?
How do these influences impinge on our assessment of the state of the English language and issues of translation? What do we do about the fact that some circles frown on generic “he”? In what way will various kinds of people understand such usage if it continues to occur in the Bible? We who are Christians, should we commit ourselves to being enriched and braced by the spiritual, moral, social, and cultural differences between Bible times and our own, or should we use language that will be most intelligible to the non-Christian world around us, and background the differences? Or can we do both at the same time? Important as these questions are, they are far too deep and complex for me to tackle them here.
But it may be worthwhile making some preliminary observations.
As I observed at the beginning of the article, a common pattern belongs to a large number of the verses whose translation is disputed. The verses involve a male human being, or else a masculine term with an associated male semantic component. But the semantic maleness occurs in a context where the verse as a whole expresses or implies a general truth applying to both men and women.
For example, with generic “he” we express a general truth: “Whoever eats my flesh and drinks my blood remains in me, and I in him” (John 6:56 NIV). “Whoever” is general. It introduces a statement that may apply in principle to any human being. Now, along with the general truth, when we use “he/his/him,” we evoke in the reader’s mind a representative individual who happens to be male. “Him” in John 6:56 pictures a sample individual. “Him” indicates that we naturally think of him as male, but the preceding context already tells us that he is a sample individual, one of many, including women. In such instances, the “gender-inclusive” translator—let us call him Jerry— eliminates the fact that the representative instance is male, while trying to retain the general principle.
Similarly, in Psalm 1:1 the “man” represents an ideal of righteousness applying to all, both men and women. So Jerry eliminates the maleness of the representative. In the case laws whose principles apply to both men and women, Jerry eliminates the fact that the representative case suggests a male.
In Matt. 7:24-27 a wise man (anēr) builds a house, and a foolish man (anēr) builds another. This parabolic picture teaches the principle that both men and women should not merely listen but receive and obey Jesus’ words. Since the principle applies to both men and women, it is only a minor matter that the parabolic picture involves two males. These males are representative of both men and women who hear Christ, so Jerry eliminates the maleness (NLT; not, however, NRSV or NIVI).
In Acts 1:21 Peter discusses the necessity of replacing Judas among the twelve apostles. He says that they should choose “one of the men (anēr).” But he also indirectly points to a larger truth, that both men and women were present when “the Lord Jesus went in and out among us” (Acts 1:21; cf. Luke 8:2-3; 23:49, 55-56). The men are representative of a larger principle of being followers and witnesses. So Jerry omits in translation the fact that Peter specified men (so NIVI and NLT, but not NRSV).
Acts 20:30 says that “even from your own number men (anēr) will arise ….” Paul warns that false teachers will arise from among the men who are elders at Ephesus. But there is indirectly a broader principle here, namely that both male and female false teachers can be expected to arise in the church (Rev. 2:20). Jerry suppresses the male marking of “men,” because the general principle can still be retained (so NRSV, NIVI, NLT).
In Acts 11:20, “men” (anēr) from Cyprus and Cyrene spread the gospel. These men exemplify the broader principle that women are also involved in spreading the gospel. Jerry speaks of “some of the believers,” dropping the male marking (NLT; not NRSV and NIVI).
In Galatians 3:26, 4:4-7, and parts of Romans 8, both men and women are denominated huios, “son/sons.” The word “son” uses the picture of a male child in order metaphorically to express a believer’s relation to the Father. Since the main point is enjoying a child-like relation to God, Jerry suppresses the male marking belonging to “son” (so NRSV, NIVI, NLT).
In Genesis 5:2 the whole race is called “man,” adam, the same word that is the name of Adam. The name designates the whole race, but uses the association with a male representative, Adam. Jerry substitutes some word like “human” or “humanity” that does not give any focus to a male connection.
The common general pattern is evident.34 At times Jerry may eliminate male marking in a parable, in a metaphor, in a case law, in a name for the human race, in a generic expression that represents a truth applying to many people. Why? Perhaps Jerry might say, “To express correctly the meaning in English, including making clear the inclusion of women.” Jerry claims, in fact, to be more faithful in translation through these innovations, because the innovations protect the intent of the original to include women.
But Jerry’s argument is very slippery. We must keep in mind several truths.
First, the original text of these passages includes women by way of “implicated meaning” or inference, rather than through the absence of male semantic marking. Second, the most accurate translation preserves, where it can, the distinction between what is expressly said and what is implicated. Third, accurate translation involves preserving every aspect of meaning of the original, not merely the “main point.” As far as possible, we include all semantic components contributed by all the words, not merely the most salient. Fourth, when we retain generic “he” and male semantic markings in translation, for the sake of accuracy, we also include the implicated meaning: the text still points to the inference that the general principles apply to women as well as men.
If he is candid, Jerry admits that the older translations were all right in their own time. “But,” he says, “the English language has changed.” What has changed? We need to become specific. There are some changes. In a former generation a sentence opening “If a man loves me,” as in John 14:23 RSV, was easily read as including women in principle. Readers today who have been trained according to the new standards may understand it as discussing a principle that would apply to men but not women. “All men” is more often understood as “all males” rather than “all human beings” (though much would depend on context). Translators may legitimately try to make adjustments in many of these cases, as the Colorado Springs Guidelines acknowledge.35 Many of these passages are not in dispute.
The ideological basis for the taboo on generic “he”
The most controversial area is the use of generic “he.” Has English changed at this point? Here the situation becomes exceedingly complicated. I cannot fathom the depths. But I can observe a good many interesting signposts.
In prestigious circles generic “he” occurs with statistically less frequency than it once did. But it does occur here and there in the secular press and in published articles and books. People’s reactions vary greatly. Some are not aware that they have encountered a “controversial usage.” They just read on contentedly. Others are aware but see no problem. Others understand generic “he,: but prefer not to use it themselves. Others understand, but with some stumbling or mild irritation. Now that attention has focused on the issue and a new standard has come into being, it seems to some that continuation in the old pattern displays insensitivity. The American Heritage Dictionary observes:
In contrast to these innovations, many writers use the masculine pronoun as generic in all cases…. This course is grammatically unexceptionable, but the writer who follows it must be prepared to incur the displeasure of readers who regard this pattern as a mark of insensitivity or gender discrimination. When a majority of writers are taking care to avoid the masculine as generic, the writer who uses it in this way may invite the inference that there is some pointed reason for referring to the representative instance as male.36
The reactions of some people may be more severe; they may be positively offended. Others feel downtrodden. A woman may say, “As a woman, I feel undervalued and left out by language that is thrusting ‘maleness” in front of me. I feel oppressed. I can’t properly deal with writing like this. If you were sensitive, you would avoid usage that has these connotations for me.”37
Translators may escape some of the odium to which other writers are exposed, because they are translators. The Bible originated centuries ago, in other languages and cultures, before the issues about English usage came to the surface. The translators might simply observe that they are faithfully rendering the Bible with respect to the original cultural context. They must not be viewed as authors who control their own meanings or modes of expression. They use generic “he” where the original used a construction with a similar function. Readers must not see “gender discrimination” where translators are simply following the meaning given to them in the original. Translators may also include in a preface an explanation of their policy.
This kind of response offers some relief, but only some. Readers may not read the preface. They may not distinguish a translation from an original work. And they may not always make in their minds the distinction between their modern context and the ancient cultural context out of which the Bible came. However, reading the Bible as a book written directly in the twentieth century is very naive. The contents of the Bible contain innumerable signs of its origin in another time and culture. These signs show readers that they cannot simply impose on the Bible modern prejudices about what is and is not a fashionable mode of expression.
The upshot is that there is indeed a change among some groups of English speakers concerning the connotations attached to generic “he.” We need to assess carefully whether we should continue to use generic “he” in the face of resistance.
But Jerry’s claims about language change do not suffice to explain fully the broader common pattern that we have observed in his translation choices. Consider, for example, the change from “son” to “child” in Jerry’s rendering of Galatians 4:4-7. At one time, “son” was a good translation in Galatians 4:4-7. What has changed? Has the meaning of huios changed? Have we discovered new facts through lexicographical analysis, showing that it lacks a male semantic component? No. Has the meaning of “son” in English changed? No, not really. What has changed, if anything, is some people’s ability to process a live metaphor involving “son” in a context in which the metaphor applies to both men and women.
How do people read the NIV, which uses the word “son”? Perhaps some people will totally misprocess the live metaphor in Galatians 4:4-7, in spite of the context of Galatians 3:26-29. They conclude, perhaps, that only men are included in Galatians 4:4-7. I do not think that this perverse misinterpretation is at all likely, but it is theoretically possible. It is far more likely that people will see that women are included, but some of them will take offense at the fact that the women are included through a metaphor that involves male marking. Or they may draw the right conclusion while feeling some fleeting sense of oddity or hesitancy. Perhaps they are just not used to the juxtaposition of women with a male-based metaphor. Or they feel a sense of irritation that women are subsumed under a male term, that some kind of “preference” or foregrounding is vaguely attached to a male.
Similarly, suppose Jerry has rendered Matthew 7 by talking about a wise “person” and a foolish “person” who each build a house. Has something changed in our knowledge of the Greek? Has the word “man” in English ceased to be able to designate a male individual in a story? Hardly. What is the problem? The story takes a male individual as an example when the example could equally easily have just taken a person as the example. Some people may stumble at the fact that maleness is imposed “unnecessarily” in two exemplary cases.
What is the problem here? Our culture has become sensitive to gender issues. Some people find problematic almost any case in which a male figure is chosen as representative of a mixed group. In their eyes “fairness” involves giving equal attention and equal prominence to females. A writer could oscillate between masculine and feminine labels, or he could constantly choose feminine labels, as with some writers’ use of generic “she.” (The repeated use of “she” could perhaps “rectify a past imbalance” that favored the masculine terms and male figures.) But one must avoid anything that appears to treat the two genders unsymmetrically, in order to avoid “insensitivity or gender discrimination.”38 One must avoid using a male as a representative instance in case laws, in the description of the righteous man, in the description of parents and ancestors, in the metaphoric description of believers’ child-like relation to God, and so on.
In other words, the basic problem, though most noticeable in the case of generic “he,” is not confined to generic “he.”39 Generic “he” appears as part of a broader cultural pattern. The culture has determined that unfairness attaches to unsymmetrical use of male and female semantic components and semantic connotations, especially in cases involving the description of mixed groups. The student on whom the professor imposes these standards asks, “Why it is unfair?” The obvious answer is that men and women are equal, and that any practice in language or society that gives unsymmetrical attention to the two shows sex discrimination. So might run a typical answer from a college professor.
In short, we are dealing here with an aspect of egalitarian and feminist ideology, pure and simple. As many acknowledge, the new standards of usage arose historically mostly through the pressure of feminist ideology. What fewer people acknowledge is that these standards can continue to be maintained only through the repeated application of this ideology when professors and cultural leaders reiterate why certain apparently innocent manners of expression are taboo. The basic reason why we tend to avoid generic “he” is that egalitarian ideology says that it is unfair.
The question of ease and familiarity
Jerry might perhaps argue that the misreading of generic “he” might occur merely because of the fact that people in the younger generation are unfamiliar with it. But this claim is implausible for several reasons.
First, as we have seen, generic “he” continues to occur, albeit at a reduced frequency, in the general press.
Second, readers encounter generic “he” when they read English literature written prior to the introduction of the new preferences.
Third, in my limited observation it seems that students in school are not taught the alleged negative connotations of generic “he,” and then forbidden to use it, until late in high school or some time in college. In earlier grades, teachers are intent on getting students to write reasonable, grammatical sentences. They just do not have the luxury of attending to generic “he.” When it finally becomes a topic for discussion, it is already too late to change the language impressions picked up in the initial stages of language learning. So now the student must be indoctrinated. He must be explicitly taught to avoid generic “he.” He must be explicitly told that it is “unacceptable” or even “evil,” even though his intuitions have not so notified him beforehand. And how will he be taught? Perhaps he will be told merely that it is not “good style.” But such an explanation, by itself, is unlikely seriously to prejudice anyone against generic “he” in the Bible. Let’s be realistic. In general, students receive some form of the explanation that I have already laid out, the explanation that says that masculine terms must not be given any unsymmetrical preference to feminine. The student thus learns that the forbidden character of generic “he” is part of a larger picture of gender and sex symmetry. That is, it is not isolated from the other forms of change that Jerry undertook to make in his translation. It is not isolated from the general structure of egalitarian ideology.
Fourth, given the way that college education pays attention to generic “he,” by forbidding its use, it is unlikely that many people will grow up totally ignorant of its use.
But let us consider this possibility. Suppose, hypothetically, someone grew up totally unfamiliar with generic “he.” It is easy to learn. Language users have a greater flexibility in adjusting to unfamiliar usages than we sometimes give them credit for. The most incredible “adjustment” is that children learn a whole language, with all its variations and peculiarities. But there are many simpler adjustments. There are languages in the world where the chief alone speaks with certain variations that the commoner does not use. Yet both chief and commoner understand one another. People quickly catch on to variations when they encounter another dialect. In previous generations millions of speakers of English learned to use “thee” and “thou” to address God. Even today, people using older liturgy and hymnody adjust to “thee” and “thou” and learn relatively quickly to make sense of it. (I am not advocating the needless retention of archaic constructions in Christian practice or in Bible translation. I am simply observing that language users have incredible ability to adjust.)
Let us consider a particularly significant case. Southern English differs from Northern English in using “you” and “you-all (y’all)” instead of just plain “you.” In Southern English “you” is singular and “you-all” is plural.40 Would not this difference cause a problem? Theoretically, the problem might be very serious. Southerners might misunderstand many occurrences of “you” in the Bible by reading these occurrences in a singular sense, though in fact a plurality of people are being addressed. What then? Do we need a “Southern version” of the Bible, with “you-all” at all the crucial points, in order to protect against gross misunderstanding?
Actually, Southerners might well feel insulted if people suggested that such a version was necessary. They might answer, “We are not dummies; we can quickly adjust to the variant use of ‘you’ when we read the Bible.”
Similarly, consider the case of generic “she.” Some writers and speakers regularly use “she” where older usage would have put a generic “he.” One might not be comfortable with this usage, but one very quickly catches on to what a writer means. Quite quickly one sees both the basic meaning and the political overtones.
Now, if a Northerner can adjust to “you-all” or a Southerner to plural “you,” if both can quickly perceive the meaning of generic “she,” what is so difficult about generic “he”? From the standpoint of fundamental, intrinsic linguistic structure, the answer is, “Nothing, really.” People have the requisite linguistic capacity to adjust and to understand.
“He” within the context of the linguistic structures of English
Since some people seem to have difficulty here, let me devote some attention to a quick, simplified analysis of some basic linguistic structures in the English language.
As twentieth century developments in structural linguistics have made clear, the meaning of any particular unit within a language system arises largely from its embedding in the larger system, where it participates in a pattern of similarities and contrasts with neighboring elements in the system.41
English like most languages has a complex pronominal system. We find personal pronouns, indefinite pronouns, interrogative pronouns, relative pronouns, and so on. These subsystems are distinct from one another, yet structurally related. The personal pronouns include singular and plural forms, first, second, and third person.42 They appear in subjective, objective, determinative genitive, independent genitive, and reflexive forms (“I/me/my/mine/myself,” “he/him/his/his/himself,” “we/us/our/ours/ourselves,” etc.). The third person singular alone has a three-way distinction of gender: “he/she/it.”
In a typical use the personal pronouns designate someone or something whose identity is specified from context. But the manner of specification may vary. Three typical modes are anaphoric, cataphoric, and deictic. The pronoun refers backwards to someone mentioned earlier in the speech or text (anaphoric); or it refers forwards to someone whose identity is specified later in the speech (cataphoric). Or it refers outwards to someone in the communicational environment (deictic). In almost all cases, the singular or plural marking and the person marking correspond to the known status of the one or ones to whom the pronoun refers. If I use “it,” I refer to a single nonpersonal item that I already know. If I use “you,” I refer to you whom I address.43
Most frequently, “he,” “she,” and “it” refer to someone or something whose identity the speaker already knows. The gender of the pronoun is then determined on the basis of semological-referential information about the referent. Thus, as we all intuitively feel, “he,” “she,” and “it” carry not only a grammatical gender marking, but, in the typical use, a semantic component indicating the sex or nonpersonal status of the referent.
Consider then the following utterance, “Sally cannot come to me unless the Father has enabled her.” The speaker who composes the sentence understands, and so does the hearer who hears. “Her” anaphorically looks back to “Sally.” “Her” is feminine because it anaphorically depends on “Sally,” and Sally is female. Similarly, “Peter cannot come to me unless the Father has enabled him.” These two sentences and many other similar sentences have meaning partly through the structural correspondences that they enjoy one to another.
Now suppose the speaker wants to generalize. “No one can come to me unless the Father has enabled ____” (John 6:65b). Because of the large-scale structural correspondences between this sentence and others, and because of the needs in the real world to express general truths, this sentence begs to be able to be completed.
Suppose now, in a hypothetical world, Peter has grown up with English but has never heard generic “he.” How will he complete the sentence? He may say, “No one can come to me unless the Father has enabled them.” The plural “them” refers back to the singular “no one.” Though such a mode of expression may seem awkward, it is not impossible. A similar use has been attested in the English language for centuries. The Oxford English Dictionary says under the entry for “they,” “2. Often used in reference to a singular noun made universal by every, any, no, etc., or applicable to one of either sex (= ‘he or she’). … He neuer forsaketh any creature vnlesse they before haue forsaken themselues” (sic from 1535).44 The logic of the English language is at work in such a usage. The use of a particular pronominal form is related to the real world referent. But in cases like these, the real world referent is in a sense a potential multitude. “Any creature” focuses on a sample “creature” using a singular grammatical form. But it uses this sample creature in order to make a statement about a plurality of creatures. The universalized form of the assertion gives it both an aspect of plurality and an aspect of singularity. These aspects occur in different dimensions, with complex relations to one another. According to a speaker’s point of view, he could treat the referent as if it were either single or multiple. By using “them” Peter has chosen to focus on the multiplicity of possible referents, rather than the singularity of the sample case, “no one.”. In our day British usage has taken over this pattern and used it much more frequently in order to avoid generic “he.”45
However, there are structural liabilities to a usage like this one. For one thing, the antecedent may become ambiguous. The NIVI contains a startling example of the problem. “By them [the ordinances of the LORD] is your servant warned; in keeping them there is great reward. Who can discern their errors? Forgive my hidden faults” (Ps. 19:11-12 NIVI; similarly NRSV). Occurrences of “they” and “them” throughout verses 10 and 11 all refer to “the ordinances of the LORD.” Verse 12 continues with “their errors,” which one would think at first would have the same reference as the earlier plurals “they/them.” Instead, the intended reference is actually to the singular “who.”
The fact that the intended antecedent is singular creates the problem. Consider again, “No one can come to me unless the Father has enabled ____.” With a slight change, we would obtain, “no one comes to me unless the Father has enabled ____.” “Comes” is the third singular form of the verb “come.” It goes with the subject “no one,” which is thereby identified as grammatically singular. In the structure of English pronominal system as a whole, there exists a pronounced preference for matching singular pronouns with singular antecedents, for obvious reasons. And “no one” is grammatically singular. The American Heritage Dictionary comments, “What is more, this solution [“they”] ignores a persistent intuition that expressions such as everyone and each student should in fact be treated as grammatically singular. Writers who are concerned about avoiding both grammatical and social problems are best advised to use coordinate forms such as his or her.”46
I would only add that the “intuition” is confirmed by a structural linguistic analysis. “Everyone likes ice cream” and “Each student sits at his own desk” both contain third person singular verbal forms, “likes” and “sits.” We are not allowed to replace them with plural verbs in order to produce “Everyone like ice cream*” and “Each student sit at his own desk.*” “People like ice cream” and “Students sit at desks” are acceptable, because they contain grammatically plural subjects, “people” and “students.” The distinction between singular and plural cuts a large swath through the grammar of the English language, so that it is impossible to excise it. “No one,” “anyone,” and “whoever” are grammatically singular.
So Peter tries again, “No one can come to me unless the Father has enabled him or her.” Peter has no solved the problem of the inconcinnity in grammatical number. And such a solution may be adequate for some uses and some contexts.47 But the result is not elegant. Structurally, it does not match the ability of other, structurally analogous sentences to use only a single noun or pronoun to do the job. Moreover, it sounds heavy, overprecise. Zinsser says of this solution, “To turn every ‘he’ into a ‘he or she,’ and every ‘his’ into a ‘his or her,’ would clog the language.”48 One may go through similar analyses of other alternatives that have been tried.49
Peter’s problem is to find a pronoun, from within the existing English pronominal system, that would refer easily and unambiguously back to “no one.” In all innocence, with no knowledge of the previous use of generic “he,” he searches. The pronoun must be a third singular personal pronoun to refer to the singular antecedent “no one.” Peter has three choices: “he,” “she,” or “it.” “It” will not do, so he has only “he” and “she.” Peter “reinvents” the usage of “he” as generic singular. Or alternatively he reinvents the use of “she” as generic singular. If, however, he knows Greek, he observes that the corresponding pronoun in the Greek of John 6:65 is masculine. That fact tips the scales in favor of using “him” in English. He has reinvented generic “he.”
Now, in a typical case, a native speaker of English does not do all this reasoning consciously. The system works unconsciously. But the system is not a rigid system, with no room for innovation, for play, for expanded usage. Thus, granted the initial hypothesis of total ignorance, the proposed expansion or “reinvention” is reasonable.
The situation is even easier for Peter if he reads a sentence already formulated by someone else. “No one can come to me unless the Father has enabled him.” The sentence goes smoothly until Peter meets “him.” Does he then stumble, because he has never seen such an odd or unheard-of barbarism? I do not think so. Unconsciously, using the language system that he has already learned, he moves as follows. “Him” is a third singular masculine personal pronoun. Look for an antecedent that is third singular. “The Father” is male singular. But if the writer intended a reflexive meaning, he would have said “himself.” So the search continues. “No one” is singular. Is it male? It is a sample including males and females. Does it work as antecedent in terms of meaning? Yes. Peter may not even notice that he has met with a construction never before seen: generic “he.” He may not notice because this use harmonizes so well with what he has already seen innumerable times, namely the anaphoric use of the third-person singular personal pronoun. Generic singular “he” is easy for Peter. He takes it in stride. It is a natural part of the structure of the English language. Without using so much structural linguistic apparatus, Zinsser observes the same thing: “But let’s face it: the English language is stuck with the generic masculine.”50
Let me make the point more vividly. Let us consider a “thought experiment,” by envisioning a science fiction scenario. Unknown to everyone, an alien superrace invades the earth. The aliens have power to remain invisible, and incredibly developed powers of mind-manipulation. The aliens remain invisible at all times. They do nothing with their power of mind control except one thing: they wipe out generic “he.” They prevent all living speakers of English from ever using an expression with generic “he” or generic “she.” They secretly alter all the books and documents from the past to eliminate generic “he.” They alter the memories of those who remember the use of generic “he.”
Once all the changes are in place, the aliens resolve not to interfere with the newborns. “Aha!” they say, “The newborns now have no chance of learning generic ‘he.’ Let us see what they will do instead.” The aliens resolve to continue their experiment for 30 years, to allow plenty of time for the newborns to grown to a mature use of English. What will happen? I venture to say that the newborns would eventually “reinvent” generic “he” or generic “she” or both.
Language learning involves generalizations of patterns. The child grows and learns the systematic difference between singulars and plurals: “dog/dogs,” “cat/cats,” “boy/boys,” “boat/boats,” etc. So the child then says “foots” as the plural of “foot.” Eventually, this mistake gets straightened out, either through someone correcting the child, or, more often, through more listening to English. The child learns that “feet” is irregularly the plural of “foot.”
Similarly, in learning a pronominal system the child learns regularities. He learns “I/me/my/mine/myself,” “we/us/our/ours/ourselves,” “he/him/his/his/himself,” etc. The aliens’ suppression of generic “he” leaves a gap in the system at the point when the child fills out the sentence, “No one can come to me unless the Father has enabled ____.” The child fills the gap using the regular pattern that he has learned. The child sees no problem.
In the case of “foots,” a real existing usage, namely “feet,” can substitute for “foots.” Hence, the child can adjust to the irregularity. But the child can never adjust to the irregularity of not having generic “he” unless there is a suitable substitute for it, to fill out the sentence “No one can come to me unless the Father had enabled ____.” As all the modern stylists have observed, and as a structural analysis of the pronominal system confirms, there is no adequate substitute. The child therefore latches onto generic “he” (or “she”) as “obvious.” He never looks back unless corrected by the punishments and lectures of others. Even if he does not latch onto this innovation himself, he would understand generic “he” the first time that he opened a Bible that the aliens had somehow overlooked and left unaltered.
What I am saying is that native speakers of English have the structural linguistic ability to understand generic “he.” Unless, in some hypothetical world, we see the triumph of a gender-ambiguous term like the proposed third person singular personal pronoun “thon,”51 or succeed in tearing up the English language to the point of total destruction or unrecognizable alteration, I do not see how anyone can abolish this ability. Friends, this ability is deep. It belongs to human nature that uses language.
We must still note some further complexities. In Peter’s world there remains a shortfall in perfect structural symmetry. Peter’s extended use of “he” cannot fully resolve the shortfall, even though his extension is based on the inherent logic of the English pronominal system. The short fall is, of course, the one with which we have wrestled from the beginning. “He” is grammatically masculine, while the referent include both men and women in the range of the sample. The range for the sample is not exactly proportional to the usual semantic load of the pronoun.
But at a microscopic level of analysis, such asymmetries occur scattered throughout language. Language users seldom even stop to notice them. For example, I have in this article used the phrase, “male semantic component.” There is a technical structural inconcinnity in this phrase, in the form of a clash in subcategorization. “Male” modifies animate nouns; it presupposes that the entity it describes is a creature with sex. But the word “component” is an abstract noun, not possessing sexuality. Readers resolve this “clash” so fast that they do not even notice it. “Male semantic component” obviously means “the semantic component ‘male,’ ” that is, a semantic feature of “maleness” attaching to some word or some part of an utterance.
There is a deep principle at work here. In many situations of real-world use of language, interpreters’ expectations about meaning drive communicative understanding in a remarkably powerful way. Even if the grammar is mangled, we frequently can make out the sense.52 I cannot produce the citation, but I remember reading somewhere that in Marcel Proust’s writings there is a complicated sentence with one too many negatives. Proust wrote the exact opposite of what he intended. But apparently for about a hundred years no one noticed it. Readers automatically grasped Proust’s real meaning and undid the effect of the extra negative!
Similarly, for centuries users of English, Greek, and Hebrew worked with masculine gender markings in their generic pronouns. Meaning triumphs over structural asymmetries in grammar. We see that John 6:65 really means “no one.”
But the lesson about the power of meaning expectations also works in reverse. If people want to find some teaching or some prejudice in the Bible, they probably will. Translators must resolve to be faithful and not to worry about every possible misunderstanding arising from prejudice.53 For the number of such possibilities is astronomical.
Reconsidering the resistance to male marking
So where are we?
In principle people possess the ability to understand generic “he.” But some do not or will not. Why not? Once we look at it this way, the answer is obvious. Propaganda fills the air, emanating from many centers of power, telling us that this use is sexist, exclusionary, oppressive. The propaganda gets under the skin even of otherwise innocent and peace-loving people. People believe it in spite of themselves. But the propaganda is just that: propaganda. It is a lie, both with respect to usage at an earlier stage in the English language, and with respect to many people who continue to this day in the earlier style of usage. It is ironic that in an age celebrating “tolerance” as a prime virtue, people should be so intolerant of linguistic variation and should slander those who deviate.
The response, then, should not be blandly to accept the status quo, but to declare the lie to be what it is, and to challenge those duped by the lie to read in a fair light not only the Bible but all English writings of previous times.
But what should we do in the interim? Some people, victimized by propaganda, will read the Bible with the wrong connotations. They will see the generic “he”; they will see the use of “son” to describe believers. If they are not too hardened, they will typically be able to grasp the basic meaning in some fashion. But they will have some struggle. They will feel, perhaps subliminally, that the Bible is subtly insensitive to women. But this subliminal message is not in fact in the Bible. “So,” Jerry says, “I eliminate generic ‘he’ in order to head off this distortion of understanding. Admittedly, I lose other nuances by pluralization, change of person, and the other kinds of restructuring that I must use to avoid generic ‘he.’ But it is worth it in order to avoid this greater and more grievous distortion.”
I would like to reply to my friend Jerry. “Yes, Jerry. I understand much of what you are saying. I understand what you are wrestling with. I understand that your motives are sincere. You desire to do good and to present the Bible as accurately as possible. I understand that you even think that you attain greater overall accuracy, in spite of having to sacrifice certain nuances.—But I do not agree with you.”
Social aspects of pronominal use
We need to reckon with one more factor. Gender in English is currently in a peculiarly complex condition precisely because people have drawn attention to gender. They have made it into a cultural theme.54 Usually we use our native tongue without conscious reflection. But people now notice generic “he” and “she” in a way that is untypical of pronouns in general. In addition, the cultural discussion has attributed to certain uses a politico-symbolic value. The use of one expression, or alternation among several expressions, symbolizes an attitude toward feminism and egalitarianism. People think that they can measure ideological progress using linguistic markers. Language police search out and destroy pockets of ideological resistance on this basis. The political dimension motivates people to look again and again, rather than release their attention in other directions. No matter what you say, you may create waves and political repercussions that you did not intend. Precisely weighed translation nuances become at this point impossible, because some readers load the translation with unwanted political overtones.
This intractable situation is exactly what Jerry thinks is his strong point. He says, “I avoid the loaded use of generic ‘he’ and generic ‘she,’ as well as ‘he or she.’ Wherever I can I use expressions that do not draw attention to themselves.” What Jerry might more accurately say is he does not intend for the expressions in his new translation to draw attention to themselves. He intends that they be free from the political fray, so that people can hear the word of God clearly. So he intends.
But unfortunately for Jerry, some people know Hebrew and Greek. Many others possess earlier translations. They make comparisons. Any time that someone reads one translation out loud and listeners have another in their hands or their memory, they can notice differences. They begin to ask questions. And then the existing cultural situation inevitably enters. “He” means what it means in contrast to “she” and “you” and so on, always withinthe context of a politico-symbolic load imposed from the culture. Likewise, Jerry’s translation means what it means in the context of contrasts with other translations, and in the context of the politico-symbolic load imposed from the culture. Cultural forces load “he” with politico-symbolic significance. The same forces immediately load with politico-symbolic significance any translation that that does somersaults to avoid generic “he.” Jerry avoids generic “he” because it is politically loaded. A significant portion of the analysts will avoid or welcome Jerry’s translation because it is just as politically loaded.55
Jerry says, “But I do not intend to be political.” Neither do the users of generic “he” intend to slight women. But given our cultural atmosphere, any move you make will be interpreted politically. I do not doubt the sincere motives of Jerry. But I doubt his foresight. And I question whether, at the level of sociolinguistic analysis, he is being evenhanded. He will not let his critics use the same logic on his translation that he used on generic “he.”56
The distortions in understanding about which Jerry worries are symptomatic of deep cultural sickness that has boiled over into elitist standards of linguistic usage. Any culture is sick if it stumbles over a story of a wise man building his house on the rock, or stumbles over Paul’s teaching that in Christ “you are a son” through the Spirit of adoption (Gal. 4:7). Such a culture is resistive, as The American Heritage Dictionary puts it, to “a particular pattern of thought.”57 It resists using a male representative to express a general truth. Many things, deep things, are needed for its healing. At the center is the gospel of Christ himself. But if there is sickness here, we do not help the sickness by sickening the Bible a little in order that the sick person can be more at home with it.
The elite of our culture have grown allergic to saying or hearing anything that might use a male term to express a larger general truth, or using language that invokes the picture of a male figure to stand for a generality, even when the context makes it perfectly evident that the maleness is subordinate to the main point. “Whoever eats my flesh and drinks my blood has eternal life, and I will raise him up at the last day” (John 6:54 NIV). What could be plainer than “whoever”? But some people are afraid to let the passage stand this way, because “him” is supposedly “sexist.”
But the language of the Bible demonstrates again and again the ethical principle that it is all right to use a male figure or a male-marked term as representative of a truth applying to both men and women. In fact, in view of the representative character of Adam as head of the whole human race, of men as heads of their families (Eph. 5:22-33), and of Christ as head of his people, it is singularly appropriate. We need not be embarrassed.
Jerry would like to quote for us Paul’s statement that “I have become all things to all men so that by all possible means I might save some” (1 Cor. 9:22 NIV). “We put no stumbling block in anyone’s path, so that our ministry will not be discredited” (2 Cor. 6:3 NIV). I agree with Jerry’s concern not to give unnecessary offense. But a translator cannot in every respect imitate Paul’s travels through the Greco-Roman world. He cannot adjust his manner of behavior for the benefit of some carefully circumscribed readership. He cannot say one thing to Jews and another to Gentiles. Once he puts out a translation, both Jews and Gentiles can pick up his translation and read it. The present state of the English-speaking cultures produces a situation where a translator is inevitably going to offend someone. There are not only feminists but also antifeminists. The antifeminists, when they compare Jerry’s translation with earlier translations, will detest it. Jerry must not write to please (an impossible task), but write to be faithful.
Jerry could respond by proposing “niche” translations. “Let us have one translation for the feminists, and another for the antifeminists. Let us have one translation for those who want everything explained and easy, and another for those who want to see the textures of the original cultural settings.”
I do think that there is room for a translation in very basic English, with very simple vocabulary, to aid young children, second language learners, and the mentally retarded.58 I think that there is room for debate about agonizing decisions that translators must make when they find no way to convey every nuance of the original. But let us look at some of the potential liabilities of adopting a policy of niche marketing on a large scale.
First, the production of different translations for different groups tends subtly to break apart the unity of Christian believers. Will Christians have to be separated into different congregations, or even different denominations, on the basis of what Bible is used in the pew? Will we have egalitarian congregations and complementarian congregations? To some extent we have them already, but translations catering to the demands of different groups increase the problem.
Second, with this policy we have no one standard Bible to memorize.
Third, we have no easy way of addressing a group that contains people from many viewpoints.
Fourth, translating for the express purpose of targeting people with an ideological viewpoint reinforces people’s tendency to regard religion as a consumer commodity. I pick off the shelf the cereal that I like. I pick off the bookstore shelf the political book whose ideological orientation I like. I pick off the shelf the Bible translation whose ideological orientation I like. The unintended message is, “We have a Bible available that will suit any ideological position.” The Bible becomes a wax nose that we may fit to any face.
Fifth, because of the pressure to make money and to maximize one’s market, Bible publishers tend to maximize their claims that their Bible version can serve everyone. A Bible consistently using gender-inclusive techniques could market itself in two ways.
This rendering of the Bible is intended for the use of those who are uncomfortable with the use of male-oriented terms to refer to mixed groups or general truths. It has sacrificed nuances of meaning for the sake of removing these offending bits, so we cannot recommend it for the general reader. But if you have trouble in this area, this book may be for you.
This Bible is for everyone. It has been translated on the basis of the most informed scholarship, and with careful attention to contemporary English style. It is suitable for private reading, devotional reading, public reading and preaching, and careful study.
Toward which description do you think publishers will naturally incline, when following fleshly desires?
But let us return to matters of substance. Jerry pointed to 1 Corinthians 9:22. I would point to 1 Corinthians 1:18-2:5, where Paul refuses to conform to the expectations of his audience, even refusing to use “eloquence or superior wisdom,” “wise and persuasive words” (1 Cor. 2:1, 4 NIV). “We have renounced secret and shameful ways; we do not use deception, nor do we distort the word of God. On the contrary, by setting forth the truth plainly we commend ourselves to every man’s conscience in the sight of God” (2 Cor. 4:2 NIV).59 What is the difference between these passages that resist cultural conformity, and the earlier passages that embrace it? Paul avoids unnecessary offense, but he will not shade the truth of the gospel. Nor will he cater to highfalutin rhetorical standards, in order to avoid the gospel’s intrinsically offensive elements, its weakness, and its grating against human pride.
Today, one of the intrinsically offensive elements is that any man, Jesus Christ in particular, could be set over us and deny us our precious “equality.” It is offensive that any message should come to us claiming authority but refusing to conform to pretentious standards for how elitist people speak and think. It is offensive that such a message should come out of backward, patriarchal cultures. We are dealing with a conflict between modern culture and the Bible. The culture says you may not use a male representative. The Bible by examples in its pages shows that you may. Since the Bible’s speech is ethically pure, we infer from the examples that the culture is wrong. The culture resists biblical claims at this point. We may not mute the claims in order to appease the culture. The difference between Jerry and me is that where Jerry sees neutral adjustment to current style, I see a subtle capitulation to one aspect of the broad cultural rebellion against the Lord and his Anointed (Psalm 2:2). We shall see what will happen (Psalm 2:9-12).
Let us be bold. Let us be bold to believe that the word of God works salvation through our misunderstandings as well as our correct understandings of it. It works salvation through our misunderstandings by provoking, jostling, undermining, and finally overthrowing them. It can do so all the better if we leave in the translations of the Bible plenty of the signs that are there in the original, signs that it is after all ethically legitimate for a male figure to represent or stand for a whole group.
Why not believe that God will use these differences between the Bible’s way of talking and that of our modern cultural elite in order subtly to rebuke and reform us, to give us life and healing and peace? Precisely at these points the Bible can enrich us, if we stand firm rather than simply caving into to what the world says is now the new standard for “offense” and “sensitivity.”
The slippery slope
Slippery slope arguments are themselves slippery. But it is worth pondering a danger here. If we follow Jerry, how far will we go with the principle of conforming to cultural sensitivities for the sake of avoiding offense? Says Jerry, “We won’t change the basic meaning.” But suppose that we can capture some nuance of meaning only at the cost of provoking bad feeling in modern readers. Then Jerry considers dropping the nuance for the sake of “greater accuracy.” “Otherwise they may misunderstand,” Jerry says. “They may incorrectly read in a discriminatory nuance.” Yes, they will. But, as I just observed, God uses misunderstanding too.
But now let us follow Jerry’s way of reasoning a little further. What about calling God “Father”? Jerry says, “Of course we translate using ‘Father,’ because that is necessary for accuracy.”
But then a voice comes disturbingly back:
Some people will be offended, Jerry. Some people will misunderstand. In our society, some people have had sinful, oppressive fathers, or no fathers at all. Some people will feel that women are excluded or slighted. Some will feel that you are claiming that God is just a male human being writ large, the ultimate chauvinist. They will think that we are saying that God belongs literally to the male sex. Or even if they do not, the feeling and connotation of it will remain beneath the surface. It is a subtle turn-off. Our culture does not associate with the word “father” the same exact things that ancient culture associated with the words pater and ’ab in Greek and Hebrew. Language and culture have changed, Jerry. We must change too. Admittedly, we lose some nuances. But we gain enormously: we gain clarity and head off all these odious misunderstandings and distortions. We may submit to a tiny loss of nuance for the sake of avoiding the big distortion, the one that will keep people from coming to God or listening to the Bible at all.
Surely the main point is that God loves us, protects us, cares for us, gives us wise guidance. Surely we do not want to claim that God is literally of the male sex. Wouldn’t we make the point even clearer, and so be more accurate, if we translated with “parent”? “Our parent, who is in heaven, let your name be precious. …”
Similar reasoning applies to the use of a masculine pronoun to refer to God or to Christ. Granted, Christ is a male human being. But why continually draw attention to his maleness, in a culture that finds this fact difficult? If we allow these concessions, others will enter from the wings, seducing us into an indefinite series of mollifications of the Bible for the sake of not “unnecessarily” offending modern readers. We cannot call God a warrior, because our culture sees war as ugly, vicious, uncivilized. We cannot call God king. “King” is male and connotes oppression under arbitrary orders. God cannot be wrathful, because it connotes that he has lost control of himself and harbors destructive emotions. God cannot threaten us with hell, because that connotes cruelty.
Then I would say to Jerry, “Jerry, you are sincere. But it is better not to overreach in translation, not to try to ‘fix’ too much, not to claim too quickly that we know what we are doing in making these changes. Do not like Uzzah attempt to steady the ark.”
We do not know all that God may do with these offensive nuances in the Bible. Nor do we know all that we will be doing if we try to head off offense. We do not know what are the full ramifications of continually finding substitutes for generic “he.” What will be the subtle effects, in the many dimensional complexities of the word of God, of language, and of culture in their interlocking? What will be the subliminal shifts? In what subtle ways will people read the Bible differently? And will it be for good or ill in the sight of God?
In many contexts, the difference between using first, second, and third person is subtle. But it touches on one of the most profound aspects of language, namely the whole process of one person communicating to another in the context of a surrounding world. The difference between using “he” and “they” may often be subtle. But again, it touches on the profound relation between the individual and the collective. The difference between a male term like “son” and a general term like “child” may be subtle. But it touches on the profundities of our sexuality, which is deep, rich, elusive, and highly mysterious. Do we know what we are doing when we make these changes?
I have spoken by way of guesswork. I do not know. I am like a child watching the patterns of waves at the seashore. I think I see some patterns, but I stand in front of a sea of vast mystery and profundity—the sea of language, used by human beings made in the image of God. Do we know fully what these changes to the new-fangled “style” signify? I doubt that anyone on this earth knows. And if we do not know, our translation should do something simple: use generic “he” when you need it to convey a singular generic meaning in the original.
1 This article overlaps at some points with my explanatory remarks on the website of the Council for Biblical Manhood and Womanhood, at www.cbmw.org. But my purposes with the two articles are substantially different. In composing this article I have profited from many people. But particular thanks are due to Dr. Wayne A. Grudem and Dr. Mark Strauss for their ideas, shared with me in unpublished papers and private correspondence, and to Don A. Carson, who shared with me a prepublication copy of his book, The Inclusive-Language Debate: A Plea for Realism (Grand Rapids: Baker, 1998).
2 Susan Olasky, “The Feminist Seduction of the Evangelical Church: Femme Fatale,” World 12/2 (March 29, 1997) 12-15; Susan Olasky, “The Battle for the Bible,” World 12/5 (April 19, 1997) 14-18; Wayne Grudem, “Comparing the two NIVs,” World 12/5 (April 19, 1997) 14-18; Joel Belz, “The Ultimate Journalistic Sin,” World 12/5 (April 19, 1997) 5; James C. Dobson, “Spooked by the Zeitgeist,” World 12/7 (May 3/10, 1997) 30; Susan Olasky, “Bailing Out of the Stealth Bible,” World 12/10 (June 14/21, 1997) 12-17; Joel Belz, “Getting Trust Back,” World 12/10 (June 14/21, 1997) 5. The Colorado Springs Guidelines concerning gender in translation appeared in their revised form in ads: “Can I Still Trust My Bible?” World 12/23 (Oct. 25, 1997) 2-3; “Can I Still Trust My Bible?” Christianity Today 41/12 (October 27, 1997) 14-15. The debate takes place in Wayne Grudem, “Do Inclusive-Language Bibles Distort Scripture? Yes,” Christianity Today 41/12 (October 27, 1997) 27-32; Grant Osborne, “Do Inclusive Language Bibles Distort Scripture? No,” Christianity Today 41/12 (October 27, 1997) 33-38; “Grudem Responds,” Christianity Today 41/12 (October 27, 1997) 39; “Osborne Responds,” Christianity Today 41/12 (October 27, 1997) 39. Other articles and an editorial in the October 27, 1997, issue of Christianity Today also touch on the problems.
3 The Guidelines appear (as revised in Oct., 1997) in “Can I Still Trust My Bible?”; they are also available through the Council on Biblical Manhood and Womanhood, P.O. Box 7337, Libertyville, IL 60048 (website www.cbmw.org). I was one of the participants in the Colorado Springs conference, and a signer of the guidelines. But I alone must bear responsibility for what is said in this article.
4 Osborne, “No,” can be taken as representative. See also Carson, Inclusive Language Debate; and Mark L. Strauss, Distorting the Bible? The Challenge of Bible Translation & Gender Accuracy (Downers Grove, IL: InterVarsity, 1998).
5 I use the phrase “male semantic component” advisedly. I am distinguishing here between grammar and semantics. The masculine gender belongs to the grammatical structure of Greek and Hebrew, as a feature attaching to nearly every occurrence of a noun or an adjective. A male semantic component belongs to the texture of the full meaning-structure, in its semological, referential, or “content” dimension. I also distinguish the particular utterance (parole) from the language system (langue). In the present case, the focus is on the particular utterance: a word occurs in a particular discourse context. For an exposition of the theoretical grounding for my point of view, see Vern S. Poythress, “A Framework for Discourse Analysis: The Components of a Discourse, from a Tagmemic Viewpoint,” Semiotica 38-3/4 (1982) 277-98; “ Vern S. Poythress, “Hierarchy in Discourse Analysis: A Revision of Tagmemics,” Semiotica 40-1/2 (1982) 107-137. An earlier article establishes the basic grammar/reference distinction (Vern S. Poythress, “Thirteen-Box Tagmemic Theory as a method for Displaying Semi-independent Language Variables,” Studies in Language 2/1  71-85). My view builds on Kenneth L. Pike’s theory of the form-meaning composite, which maintains a distinction between grammatical and referential dimensions, but not a strict separation of the two in the occurrence of particular linguistic units (Pike, Language in Relation to a Unified Theory of the Structure of Human Behavior [2d ed.; The Hague/Paris: Mouton, 1967], especially 62-63). Pike’s view, supported by both the linguistic and the philosophico-theological sphere, has subtle implications for the manner in which I develop this article. But I need not elaborate at this point on the nature or integrity of my theoretical foundations.
1. The generic use of “he, him, his, himself” should be employed to translate generic third-person masculine singular pronouns in Hebrew, Aramaic, and Greek. However, substantival participles such as ho pisteuōn can often be rendered in inclusive ways, such as “the one who believes” rather than “he who believes.”
2. Person and number should be retained in translation so that singulars are not changed to plurals and third-person statements are not changed to second- or first-person statements, with only rare exceptions required in unusual cases.
Throughout my discussion of generic “he,” I focus on the anaphoric form, linked to an earlier generic term like “anyone” or “whoever.” The Colorado Springs Guidelines allow “he who …” to be changed to “the one who …,” so such changes are not in dispute.
9 “We” and “they” also have generic uses. “Although used generically, these personal pronouns we, you, and they retain something of the specific meaning associated with the 1st, 2nd, and 3rd persons respectively. They are therefore not wholly interchangeable” (Randolph Quirk, et al., A Comprehensive Grammar of the English Language [London/New York: Longman, 1985] 6.21).
10 Ernst-August Gutt, Relevance Theory: A Guide to Successful Communication in Translation (Dallas: Summer Institute of Linguistics, 1992), 72. See also Ernst-August Gutt, Translation and Relevance: Cognition and Context (Oxford/Cambridge, MA: Blackwell, 1991). The implications are many. The rather loose understandings of “dynamic equivalence” principles for translation that I sometimes hear bandied about today must undergo serious modification. We need careful qualification, restriction, and nuancing. Language and translation are excruciatingly complex.
11 I am not equating generic singulars belonging to different languages. The structures are different. I am simply observing that as a matter of fact, in the cases about which we are talking, there is considerable overlap between the meaning-functions of generic singulars in two languages.
12 Grudem, “Yes,” 32, furnishes examples, as well as a quote from The Associated Press Stylebook and Libel Manual (1994), “use the pronoun his when an indefinite antecedent may be male or female; A report attempts to protect his sources.” (Not his or hersources.)”
14 The composer of Psalm 109 may have had one particular male enemy in mind (Shimei?). But by inclusion in the Psalter the psalm receives a generalizing or generic sense. “He/him” is representative of an enemy that the godly may have, and it is fulfilled in a uniquely concentrated way in Judas Iscariot (Acts 1:20).
19 But when abstract or nonanimate nouns become personified, the situation may be different. Moreover, in Hebrew certain kinds of ideas tend to have feminine nouns (GKC §122). As usual, there are lots of complexities.
22 U.S. News & World Report, Nov. 27, 1995, p. 18; ibid.; Chicago Tribune, May 12, 1997, p. 1. I am grateful to Dr. Wayne Grudem for supplying me with these instances. Note that when “man” is used as an expression for the human race, it always appears in the singular without an article. This feature clearly differentiates it from uses like “a man” and “men.”
28 Johannes P. Louw and Eugene A. Nida, Greek-English Lexicon of the New Testament, 9.1, allows that anēr has the meaning “person, human being, individual” in Matt. 14:35 and a number of other cases. That a man can be a representative of a general truth appears from Psalm 1 and Psalm 32:2. But, as we have seen in discussing ’ish, this observation does not eliminate the existence of a male semantic component in the representative, the single person who embodies the general truth. BAG also acknowledges the use of anēr for “someone.” But again, inspection of the actual data shows that in many cases the “someone” is in fact male or is a representative figure (still male?) expressing a general truth. Likewise, the information presented in Horst Balz and Gerhard Schneider, eds., Exegetical Dictionary of the New Testament (Grand Rapids: Eerdmans, 1990), 1:99, needs critical sifting.
32 If the target language did not contain two words adequately representing the difference between “king” and “master,” we would, of course, be forced to compromise. But we are not occupied at the moment with this problem.
34 It would deflect from my main argument to observe that, even though there is a common pattern, the different cases are far from being on the same level. In some cases, as in Acts 1:21 and 20:30, the texts are directly addressing a specific historical situation, and the broader general principles lie far in the background. We clearly need a translation with “men.” At the other extreme are cases like “If a man …” in John 14:23 (RSV). In current English “a man” in a typical sentence not only carries a male semantic component, but suggests to many exclusive maleness. Because of broader cultural expectations, people have a firm expectation that male marking will not at all be used in conjunction with the expression of a generic truth. Hence, the Colorado Springs Guidelines explicitly permit the rendering, “If anyone …” in places like John 14:23 (see guidelines A6, A7, A8).
The case laws are difficult. The special genre, in addition to the plainly exemplary character of the descriptions, gives evidence of their universality. The form, one suspects, depends as much on a convention in judicial writing as on a generically expandable function of ’ish. If we translate using “a man,” we make visible more of the flavor of this judicial style. But, in spite of the context, some modern readers will have in their minds the exclusion of women. Given the vexing situation in English-speaking culture, there is no way of doing it all. Despite opponents’ impressions to the contrary, the Colorado Springs Guidelines allow discretion in difficult cases, through qualifying expressions like “ordinarily,” “unusual exceptions in certain contexts,” and “some details may need further refinement.”
Generic “he” poses a special problem. As long as a preceding “anyone,” “everyone,” “whoever” or a noun phrase indicates the scope of application, readers can take a subsequent “he” in stride. The problem then, as we shall see, is not so much mistaking the meaning as being offended by the style.
37 Such expressions of distress ought, of course, to be taken seriously. But there is a deeper side to some of them. Our culture has learned the language of “victimization.” It has learned to manipulate people and to push them into action by putting them in a position where refusal would allegedly show heartlessness. This kind of manipulation has become so common that many may use it “sincerely,” that is, without consciously realizing that they are manipulating through deceit and distortion. They have absorbed the distortions unconsciously. They have talked themselves into a certain view of their situation, or have been talked into it by others. Through extensive exposure to our culture, playing the victim has become second nature to them.
39 For further exposition of the relation of generic “he” to a larger sociolinguistic picture, see Valerie Becker Makkai, “Correctness in Language: Political and Otherwise,” 1996 Presidential Address, The Twenty-Third LACUS Forum 1996, ed. Alan K. Melby (Chapel Hill, NC: Linguistic Association of Canada and the United States, 1997), 3-25.
41 The idea of meaning through systemic contrast goes back to Ferdinand de Saussure, Course in General Linguistics (rev. ed.; Glasgow: Fontana/Collins, 1974). But it is displayed again and again across the pages of modern linguistics textbooks. I would note, however, that the system of language is not an isolated, independent entity, but one intelligible only within an embedding as one subsystem with a framework of world and culture within which real people live, think, and behave—all superintended by God. See Kenneth L. Pike, Language; Pike, Linguistic Concepts: An Introduction to Tagmemics (Lincoln: University of Nebraska Press, 1982); Pike, Talk, Thought, and Thing: The Emic Road toward Conscious Knowledge (Dallas: Summer Institute of Linguistics, 1993).
44 The Oxford English Dictionary, ed. J. A. Simpson and E.S.C. Weiner (2d ed.; Oxford: Clarendon, 1989) 17:928. See Dennis Baron, Grammar and Gender (New Haven/London: Yale University Press, 1986), 191-97.
47 But not for Bible translation. “He or she” draws attention to itself, and also suggests that the authors in the ancient context were making a conscious effort to be explicit in including females. Once again, implicated information in the original has becomeexplicit information in the translation.
51 See Zinsser’s sarcastic dismissal in ibid., 120. Even if English could be altered so as to possess the extra pronoun “thon,” I would guess on the basis of the arguments already given that native speakers would still be able to understand generic “he” and generic “she,” though they would see in such a use a focus being given to a male or female sample out of the general case.
52 One might also note the situation in pidgin languages. The grammar may be very simple. Communication succeeds in many instances because the meaning expected from context powerfully guides the interpreter.
53 I agree with the importance of checking translations with native speakers. But we focus on clearing out unintended ambiguities or failures of communication that may arise even for the spiritually sympathetic reader; we ought not to appease the hostile reader.
55 This reaction need not involve only Christians. Perhaps some pagans, on hearing the news of gender-inclusive translation, will rejoice that Christians themselves are finally admitting how outmoded the nonegalitarian thinking of the Bible is.
56 If I may exaggerate: Jerry expects antifeminists to understand Jerry’s pure motives, to isolate his translation style from all political meanings, and to accept with equanimity the losses in meaning-nuance that ensue. On the other hand, he expects from feminists only obtuseness, intolerance, and crass attribution of political meanings if he uses generic “he.” Antifeminists should thank Jerry for his high compliment to their angelic insight, patience, and gracious disposition. But they can also ask him to be more realistic about themselves, the feminists, and the larger cultural configuration. And they might gently suggest that he might be more realistic about himself. Is he tempted to pay disproportionate attention to the concerns of the loudest cultural voices? Does he have too much confidence in his own superiority, leading him to think that he can translate successfully, without being caught in the same politicized environment under which ordinary mortals labor?
From here on the argument becomes so complex that I find it impossible to follow all the turnings. For instance, Jerry claims that the political effects are less in one direction than in the other. Generic “he” is immediately visible to language police if it occurs even once. By contrast, the practice of avoiding generic “he” is only indirectly detectable by comparing the patterns of translations over many verses.
But what are the long-range effects? If a translator retains generic “he” wherever he needs it, the shocks to feminist readers decrease with the number of occurrences. Gradually readers adjust to recognize that no slight to women is intended. On the other hand, if the translator avoids generic “he” throughout, the shocks to antifeminist readers, as well as to many serious Bible students, increase as the number of affected verses mounts into the thousands. People may think, “Look how much the translator was willing to sacrifice on the altar of cultural respectability. The price is too high.” The result may be long-range loss of trust.
There is no translational solution that will raise us above the political fray.
58 Such a simplified version should, however, be clearly labeled, “The Bible Rendered in Basic English,” or “The Bible in Simplified Language,” or the like. The preface should carefully and honestly explain its purposes and limitations, to avoid giving readers a false impression. Note what I say below concerning marketing.