This continues the series from Part 1 and Part 2. In the former I started to examine Joel Hoffman’s claim that the Greek word anthropos has a male meaning component. In the latter I introduced the concept of markedness and outlined how it might be applied to gender in Greek. Now I want to bring these two together by considering how the markedness helps to explain how anthropos is used.

As I explained before, anthropos is one of a group of Greek nouns which can be either masculine or feminine. The technical word for this is “epicene”. The feminine form of anthropos is rare, and not found in the New Testament. I guess many epicene nouns are much rarer in the feminine than the masculine – although the opposite is true, at least in the NT, of parthenos “virgin” which is usually feminine, but presumed (on the basis of usage elsewhere – the gender is not marked in the NT text) to be masculine in Revelation 14:4 (referring to men only) and perhaps 1 Corinthians 7:25 (referring to both men and women).

Let us now consider how anthropos is used in the New Testament. According to the rough figures in my Modern Concordance to the New Testament (Darton, Longman & Todd 1976) the 552 occurrences can be divided as follows: 88 in the phrase “son of man”, mostly referring to Jesus but including Hebrews 2:6 which I discussed here yesterday; 5 referring to Adam; the 5 occurrences I discussed earlier in which there is a contrast with “woman”; 29 other cases referring to Jesus; 39 referring to other named individuals; 43 referring to unnamed individuals; 5 referring to inhabitants or citizens; 1o referring to the self or nature; 2 in the phrase “man of God”; and the rest, more than 300, referring to “MAN, HUMAN(ITY) – PEOPLE, EVERYBODY, EVERYONE – SOMEONE, ANYBODY”.

Some of the 111 references listed as to named or unnamed individuals are in fact to groups which may well have included women. But it appears that every reference to a single individual is to a man, an adult male, rather than a woman or a child. I say “appears” because in many cases the gender and age of the referent is not otherwise stated. But I would not dispute a claim that in the New Testament anthropos never refers to a specific woman – although it does (with feminine gender) in other Greek literature.

Here is the Greek text of the most convincing example of feminine anthropos, from Herodotus 1:60, the first of the six examples Suzanne quotes in English translation:

οἱ ἐν τῷ ἄστεϊ πειθόμενοι τὴν γυναῖκα εἶναι αὐτὴν τὴν θεὸν προσεύχοντό τε τὴν ἄνθρωπον καὶ ἐδέκοντο Πεισίστρατον.

Here we have two feminine epicene nouns: he theos “goddess”; and he anthropos, rendered “human creature” and referring to the woman who had just been described, more typically, as gune. Presumably she is called anthropos because she is being contrasted with the goddess Athene. I can’t help thinking that anthropos would have been used in a somewhat similar incident in the New Testament, at Acts 12:22, even if the referent had been a woman.

To complete this study it is important to look at that majority of the occurrences of anthropos classified as referring to “MAN, HUMAN(ITY) – PEOPLE, EVERYBODY, EVERYONE – SOMEONE, ANYBODY”. The significant point here is that only a very few of these more than 300 refer to men rather than women or indeed have any gender significance at all. Many of these gender generic examples are plural, but there are also a considerable number which are singular but still gender generic. Consider for example the use of singular anthropos in Mark 7:14-23 and Romans 3:28, teaching which surely applies to women as much as to men. So there really is no evidence to support the common claim that anthropos is gender generic in the plural but specific to men in the singular.

To put this in the language of markedness, this very common use of anthropos to refer to men and women indefinitely without specifying gender seems to show that this word is the default or unmarked one for referring to human beings in general, singly or in groups. In the thinking of the time (and to some extent today) the default or unmarked person, the prototypical person (to use the language of another semantic model), was a man, an adult male.

This explains why when it was known that a person was a woman or a child it was normal to use not anthropos but a different word, one marked as referring to a female or a young person. But in exceptional cases, perhaps for stylistic variation or to contrast with a god or an animal, anthropos could be used of a specific woman or child. And when used of a woman it was also marked as such with its gender, feminine rather than the unmarked masculine. But the selection of grammatical gender seems to have been independent of the choice between anthropos and other words.

The implication of this is that anthropos is a word entirely devoid of gender marking within itself, that it in no sense “means” “man” to the exclusion of woman. The fact that it is rarely used of specific women is entirely explained by markedness theory, and does not indicate any male meaning component.

I accept that this is only the outline of an argument, which would need to be firmed up by a more careful examination of the evidence, not relying on the sometimes questionable classifications in my concordance and not restricted to the New Testament. But I think I have given enough evidence to show that Joel’s hypothesis that “one meaning of anthropos is “man”” is unlikely to be correct.

In Part 1 of this series I outlined the issue with the Greek word anthropos and Joel Hoffman’s claim that this word has a male meaning component. I showed that the word is used only extremely rarely to contrast with words meaning “woman”. I also linked to evidence that the word is used of specific women, although rarely and never in the New Testament, in which case it has feminine grammatical gender.

At this point I want to introduce my readers to the concept of markedness. This is an important idea in linguistics, one which was first developed in the field of phonology but is now proving useful in describing other aspects of languages.

Steve Runge helped me to learn more about markedness in a series of posts earlier this year on his NT Discourse blog. The most important post for this discussion is the first one; in his other with the markedness tag he applies this theory to Greek verbs and discourse. Here is part of Steve’s introduction:

Markedness is an organizational framework for taking a complex set of data and organizing it into meaningful groupings to facilitate description of the members. The organization is accomplished by taking the most simple, basic member of the set, and using it as the canon against which the other members are contrasted. The most basic member is referred to as the default. The other members of the set are then described by how each differs from the default and from the other members. The default option is the one used when, to paraphrase the Hallmark commercial, “you do not care enough to send the very best.” In other words, when there is nothing special that one wants to signal as present, the default is the natural choice. For this reason, the default is generally the most frequently occurring member of the set. It does not signal or mark the presence of any special feature. In this way the default is also called the unmarked option.

This system that I have described is an asymmetrical approach to markedness, where each different member of the set marks the presence of a different, unique feature. There is another approach to markedness that is far more widely known within NT studies, though it is not used nearly so prevalently in linguistics proper. I mention it here for clarity sake, knowing that it may cause confusion for some. The intention is to show what I do not mean by markedness when I use the term.

The second approach to markedness is the symmetrical one. …

At this point I would like to make it clear that I am also using the “asymmetrical approach to markedness”, and not “the symmetrical one”. Read Steve’s post for more explanation of the difference.

I would like to apply this concept of markedness not to Greek verbs and discourse but to Greek gender and gender-related words. Of course I can only do this very inadequately in the course of a short blog post – there is very likely enough material here for a PhD. But I would like to make some provisional observations based on my experience of how Greek works.

First, I think Greek makes a clear distinction between gendered lexical items and grammatical gender. At the lexical level one can distinguish between gender pairs of words which are very different in form, e.g. ho aner “man” and he gune “woman”; pairs which differ only in their ending, in effect declining like adjectives, e.g. ho adelphos “brother” and he adelphe “sister”; and words which can be masculine or feminine depending on the gender of their referent but do not change their form, e.g. ho diakonos “servant/deacon (male)” and he diakonos “servant/deacon (female)”. The word we are mainly discussing here, anthropos, fits in the third category, although its feminine form is rare and not found in the New Testament. Some words fit in more than one category: the female form of ho theos “god” can be he theos, as in Acts 19:37, or he thea, in Acts 19:27.

In Greek, as in all gendered languages as far as I know (certainly also in French, Latin, German, Italian, Russian and Hebrew), the masculine plural is regularly used to refer to groups of mixed gender. This is already a strong indicator that masculine is the unmarked or default gender and the feminine gender is marked. Further evidence of this, at least in Greek, comes from the regular use of the masculine gender in indefinite sentences, e.g. to refer back to the genderless pronoun tis “someone”/”who”, even when the sentence is clearly applicable equally to men and women.

So how might these principles apply more specifically to anthropos? I will discuss that in Part 3.

I’m sorry that this post is rather technical, and so may be hard for some of you my readers to understand. But in view of some of what I have read recently on blogs it is important to get these matters right.

There has been quite a lot of discussion on various blogs about whether the Greek word anthropos “means” ” man”, in any way to the exclusion of women. In particular Joel Hoffman has taken the position, here and here (see also this post), that

one meaning of anthropos is “man,”

and that in some places where the word is used

the Greek text means to emphasize “man” over “woman” … anthropos emphasizes “man” in contrast to “woman.”

I have strongly opposed Joel’s position in comments on these posts. Also disagreeing with Joel have been Suzanne McCarthy, here and here as well as in comments on Joel’s blog, and Kurk Gayle, here with links to several other related posts, also here and in comments elsewhere. See also Mike Aubrey’s related post, here. I also touched on this issue in two previous posts of mine, here and here. Read the comments on each of these posts.

In particular, I had to correct Joel for the following demonstrably false statement which he made in comment 5 here:

3. I still haven’t seen any convincing evidence from extrabiblical sources to support anthropos referring specifically to a woman. Did I miss one?

There is convincing evidence, provided by Suzanne, and Joel had earlier commented on this post showing that he had seen the evidence. I have twice asked Joel to correct this error. His response the first time suggests that he may have intended to qualify his statement with “with the masculine determiner”. With this qualification the statement would be correct: when anthropos is used of a specific woman or a group of only women it is grammatically feminine. But Joel has refused to correct or withdraw his original statement, which, without the qualification and so stating that anthropos is never used of specific women, is factually incorrect and highly misleading.

However, my main purpose here is not to correct Joel on a detail, but to look in more detail into why his overall approach to this issue is wrong-headed.

The essential feature of Joel’s argument seems to be this: because anthropos is sometimes used to contrast with words meaning “woman”, that implies that there is something male about its essential or core meaning. I consider this to be an incorrect deduction. The core meaning of a word is not found by looking at a few unusual examples.

This usage of anthropos in contrast to a word meaning “woman” is in fact rather rare in the New Testament. According to my Modern Concordance to the New Testament (Darton, Longman & Todd 1976), out of 552 occurrences of anthropos in the NT only five are “IN RELATION TO WOMEN”. Three of these, Matthew 19:5, Mark 10:7 and Ephesians 5:31, are direct quotations from the LXX Greek translation of Genesis 2:24, rendering Hebrew ish. So this is translation Greek – and as linguists know it is never good practice to study the characteristics of a language from a translated text.  A fourth case, Matthew 19:10, immediately follows one of these quotations and so can be understood as an echo of the translation Greek.

This leaves just one example, 1 Corinthians 7:1. I dealt with this issue as long ago as 1988 (long before I had a particular interest in gender issues), in an essay which I posted on this blog in 2006, in a section dealing with possible quotations in 1 Corinthians from a letter to Paul from the Corinthians:

A second characteristic is the use of ἄνθρωπος [person (anthropos)] in both 7:1 and 7:26 for man as opposed to woman, where ἀνήρ [man (aner)] is normally expected. These are the only unambiguous examples in Pauline writing of this use, except in Ephesians 5:31 where Genesis 2:24 is quoted. This provides added evidence that there is a quotation in 6:18, for in context the ἄνθρωπος in this verse is probably male. The similar use of ἄνθρωπος in 7:7, contrasting with the regular pairing of references to men and women in 7:1-16, strongly suggests that here also there is an adapted quotation from the Corinthians: θέλω δὲ πάντας ἀνθρώπους εἶναι ὡς καὶ ἐμαυτόν [but I want all people (anthropos) to be as also myself]. The similarity of this to καλὸν ἀνθρώπῳ τὸ οὕτως εἶναι [good for a person (anthropos) to be like this] in the acknowledged quotation of 7:26 is more evidence for this further quotation.

In other words, I am suggesting that this non-generic use of anthropos was a characteristic of the letter from the Corinthians, reflecting the dialect or idiolect of its author. It certainly doesn’t seem to be characteristic of the rest of the New Testament.

So we have effectively shown that this gender specific use of anthropos is extremely rare in the New Testament, being found only in translation Greek and in an exceptional case. This is already enough to cast serious doubt on the proposition that gender is a core component of the meaning of the word. But I accept that further proof is needed.

There is still a lot more that I would like to write about this issue, but this post is getting too long already, so I will continue in Part 2 and Part 3.

A few days ago I posted on The value of men, women and sheep. The issue I brought up there, the meaning and translation of Matthew 12:12, generated quite a lot of discussion on this blog and elsewhere. Suzanne, in a comment on one of Joel Hoffman’s posts, raised the issue of the rather similar passage in Luke 13:

Jesus Heals a Crippled Woman on the Sabbath

¹⁰ On a Sabbath Jesus was teaching in one of the synagogues, ¹¹ and a woman was there who had been crippled by a spirit for eighteen years. She was bent over and could not straighten up at all. ¹² When Jesus saw her, he called her forward and said to her, “Woman, ^[a] you are set free from your infirmity.” ¹³ Then he put his hands on her, and immediately she straightened up and praised God.

¹⁴ Indignant because Jesus had healed on the Sabbath, the synagogue leader said to the people, “There are six days for work. So come and be healed on those days, not on the Sabbath.”

¹⁵ The Lord answered him, “You hypocrites! Doesn’t each of you on the Sabbath untie your ox or donkey from the stall and lead it out to give it water? ¹⁶ Then should not this woman, a daughter of Abraham, whom Satan has kept bound for eighteen long years, be set free on the Sabbath day from what bound her?”

¹⁷ When he said this, all his opponents were humiliated, but the people were delighted with all the wonderful things he was doing.

1. 12 The Greek for Woman does not denote any disrespect.

Luke 13:10-17 (TNIV)

Structurally this story is rather similar to Matthew 12:9-14, although here the healing takes place before the objection. Also the comparison of the value of a human being to that of an animal is only implicit in this story, but explicit in Matthew 12:12. However, the two passages cannot be considered parallel passages. That is because the Matthew passage has much closer parallels in Mark 3:1-6 and Luke 6:6-11 – although only Matthew makes the human-animal comparison explicit.

Interestingly in Luke’s account of the man with the withered hand he is initially introduced as anthropos, i.e. his gender is not specified, but later referred to as aner, confirming that he is male. This may be simply stylistic variation, but from memory it is not the only place where Luke has aner and other gospels have anthropos.

But Luke balances his story of a man being healed in the synagogue with the rather similar story of a crippled woman being healed in the same setting, which has no parallels in the other gospels. Clearly Jesus showed no gender discrimination in his healing, and Luke wants to make that clear.

I did find one interesting translation point in the story of the crippled woman. In TNIV, and most other translations (including NLT and CEV, but not The Message), she is implicitly compared with an ox or a donkey, and considered to be of much greater value.

Now the Greek words used for these two animals, bous and onos respectively, work like anthropos in that they do not specify the gender of the animal. All these words can be grammatically feminine (without changing the form of the noun) when referring to a specific female: onos is feminine in Matthew 21:2,7, but masculine in Luke 13:15; bous is also masculine here and, like anthropos, in every case in the New Testament where its gender is specified. However, again like anthropos, when used in generalisations onos and bous are masculine, but this should not be taken as specifying the gender of the animal.

Thus in Luke 13:15 the translation of onos as “donkey” is correct, as this word is gender generic in English. But the rendering of bous as “ox” is more questionable. First we need to clarify what is meant by “ox”. Here is the definition from answers.com:

ox (ŏks)

n., pl. ox·en (ŏk‘sən).

1. An adult castrated bull of the genus Bos, especially B. taurus, used chiefly as a draft animal.
2. A bovine mammal.

[Middle English, from Old English oxa.]

Which definition did the TNIV translators have in mind? If the latter, then they are exegetically correct. But would it be normal language, among their target group of speakers of contemporary English, to refer to bulls and cows as oxen? I think not. The generic term in current use is “cattle” or “bovine”, or colloquially “cows” although as a country boy I find it strange, and a sign of city-dwellers’ ignorance, when male cattle are called cows.

I would suggest that the word “ox” is more or less obsolete in modern English. If it is used at all, it is used in the rather specific sense 1 I quoted above, of an adult castrated male bovine. This is probably how it is understood by most Bible readers today who have any real understanding of the word.

So, I would suggest that TNIV, while being carefully gender generic about humans, has failed to follow the same principle when referring to cattle. To be consistent, it should drop “ox” in this verse and 14:5 and find an alternative gender generic rendering.

“Ox” can remain in 14:19, 1 Corinthians 9:9 and 1 Timothy 5:18 as these verses are about draft animals which in the cultural context probably would have been castrated males. But the cultural context in 13:15 is quite different. Here we have a person, probably an ordinary villager, who keeps one bovine and one donkey in a shed or courtyard and has to untie them to lead them to water. This is still common practice in the Middle East. And the bovine would normally have been a cow, a female – such families would have had little use for an ox but a cow would have been an important source of nutrition.

So in 13:15 I would actually suggest “cow”, as in The Message. I thought of “one of your cattle”, but that implies that the person has more than one whereas the text and the cultural context suggest the opposite. Thus “cow” is better. If anyone complains that “cow” is not gender generic, one can point them at this sense in the dictionary:

A domesticated bovine of either sex or any age.

But if a choice has to be made between a term with male connotations and one with female connotations, in this case of a comparison with a woman surely the more female word is preferable.

I hope no one thinks this is political correctness gone mad! My point is simply that translations need to be accurate. To use a gender specific term to translate a gender generic one, whether for humans or for animals, is inaccurate translation.

The ironic thing here is that in first century Palestine a cow, a female producing milk, was probably more valuable than an ox, a castrated male – at least to an ordinary village person. Jesus in his parable compares the woman to a cow, more valuable than an ox, or for that matter than the sheep of Matthew 12:12. Translators should not demean women by comparing them to something less than Jesus had in mind.