Translating Punctuation when there is No Punctuation to Translate

Jonathan Morgan, on our share page, asks this,

One thing I have heard a number of times is the assertion that “Greek has no punctuation”, and that as a result we can choose to repunctuate the *English* in any way we like, because “it’s all just been added by the translator anyway”. I’ve never been entirely convinced by this…

First, good for you not being convinced by the apparent, and incorrect, logic of “no punctuation in the original means we can punctuate the translation any way we like.” We are never free to translate “any way we like.” The goal is accuracy. Secondly, there’s an underlying assumption (if I myself may assume such) in the “logic” that punctuating is not translation. The use of punctuation in the destination text most certainly is translation as is such things as paragraph breaks and section breaks.

English uses punctuation. So, punctuation is required in the translation, or it wouldn’t be clear and natural—it wouldn’t communicate to an English audience. However, just because there were no punctuation marks, per se in the original, does not mean the function of punctuation was not performed in the original. The function of punctuation is to generate meaning pauses for the reader so as to generate cognitive chunking (think of this as taking bites of the text with your mind). And so it is such a basic cognitive requirement that, as far as language goes, this function is a language universal. So, the function is there; we just need to determine how that function is formally captured in the original so we can accurately translate the meaning into a language that uses punctuation marks.

Before I give some explanation, I’ll point out that the web page you point us to gives a good explanation. The question the web page answers shows a wrong assumption about the translated text. It says, “Holman, CEV and others place the comma in a way that implies that Jesus had already risen, before the first day of the week,” citing Mark 16:9–“very early on the first day of the week, after Jesus had risen to life, he appeared…” While the translation might imply that Jesus had risen before the first day, the translation does not say that. It simply and only says that the resurrection happened before the appearance, and that Mary saw Jesus very early. Sometimes I think we judge a text by the cover we ourselves project on to it. While an important criteria for translation is to be unambiguous, we can’t prevent people from wrongly interpreting a text no matter how clearly we write it (see 2 Peter 3:15-16). I think there’s a tacit contract between translator and reader that each will do the best they can. There are no major translation publications where the translators have intended to lead the reader astray. I felt I had to get this out of the way.

I’m going to illustrate from the Greek. I assume Hebrew and Aramaic are analogous. Basically, the question is: What are some of the mechanisms ancient koine Greek used to “punctuate” the Biblical text?

Well, for example, Mark (and others) frequently used καί (KAI, ‘and’) to mark a sentence break.[1] Open an NASB to Mark 3:13-20 for a good illustration of this. The function καί brings to the text is to mark the closing and opening of two sentences. This “punctuation mark” (if you will) is much like our English period and a capital letter. Δέ (DE, ‘and’, ‘so’) frequently performs the same function.

Also, one should not think that the Gospel of Mark is rapid fire because he uses so many καί–“and this, and this, and this”. That’s not what is going on. That’s interpreting the Greek using an English idiom (ie. way of thinking with our language). Many times καί “provides” the punctuation between two sentences.

However, let me be clear here. Καί and δέ perform other functions, too; the ones we normally think of them doing. Καί connects two semantic items which are otherwise equal. Δέ adds supporting material to what has just been written. However, just like so many things in translation, there is no one-to-one mapping between the form in the original and its analog in the destination. The mapping between the languages is nearly always many-to-many. That is, the characteristics that a specific form brings to the text in the original will map to multiple forms in the destination and vice-versa.

This complexity is why the Tower of Babel was so successful, and it makes translation hard. I’ll also point out that translating punctuation is clearly one place where a naive adherence to a formal equivalent methodology breaks down. A naive adherence that no formally equivalent translation follows. Since there was no punctuation in the original, there’s no way to formally map it to the destination. The point being: Even the formal equivalent methodology must follow a functional equivalent methodology when it comes to punctuation.

So, there were no punctuation marks in the original; but that function is dispersed through many Greek forms. And one of the characteristics of those original forms (a punctuation function) maps to the many punctuation marks in English. So, it’s not arbitrary. But, nor is it formally equivalent.

Furthermore, Greek has flexible word order, but it is certainly common for the Greek sentence to either begin or end with a verb. This, too, tends to mark the breaks between sentences. Obviously, I’m not describing this in a mechanically precise way. Nor is its use or non-use determinative. To illustrate, I’m saying that the sentence in Acts 1:2 ends in a verb and the one in Acts 1:3 begins with one.

ἄχρι ἧς ἡμέρας … ἀνελήμφθη. οἷς καὶ παρέστησεν ἑαυτὸν ζῶντα…
“Until which day…he was taken up. To whom he also presented himself alive…

It’s very natural and expected to have the phrase οἷς καὶ pre-positional to the verb and still think of the verb as being “first” in the sentence. An author will vary the verb’s position for a variety of reasons. I believe “punctuation” to be one of those reasons. Again, there’s no, neat, sweat, simple one-to-one mapping.

There are other forms, too. I may be wrong, but I’ve often thought that one way of making direct speech very clear is the often used combination of two verbs of speaking used in close proximity. For example, ἀποκριθεὶς ὁ Ἰησοῦς εἶπεν αὐτοῖς (Literally: “Answering, the Jesus said to-them.”) In English, we punctuate with double-quotes. In Greek, the ἀποκρίνομαι does more than just help fulfill this punctuation function, it also characterizes the way Jesus said what he said. Again, it’s many-to-many. An accurate translation is: Jesus answered, “…” or even Jesus responded, “…”. For our purposes here, note the quotation marks in the translation. They are not in the original, per se. But, their function is.

There’s much more that could be said. Hopefully, this provides enough meat so you and others will have confidence that punctuation is not arbitrarily decided. Punctuation, like every other form (or symbol as used in semiotics) signals something. The way at getting at that “signaling something” is to ask and answer, what function is it performing. Since the function punctuation performs is so cognitively basic, we expect the function to be in the original even when the English way of performing that function is no where to be found. I hope my start of an answer generates some further examples in the comments as well as some discussion.

[1] The so called definition that καί and δέ mean ‘and’ or ‘but’ is far too simplistic, and it is either wrong or at best an insufficient explanation. The continuity or discontinuity provided by the English ‘and’ or ‘but’ is provided in the Greek by the semantics of the sentence. Καί connects two equal items; δέ adds supporting material. Again, there’s a mapping between the original and the destination languages, but one cannot simply match the forms.

6 thoughts on “Translating Punctuation when there is No Punctuation to Translate

  1. Dannii says:

    A related question:

    I suspect most people translate from Hebrew/Greek Bibles which do actually have punctuation. How would the editors of those texts have introduced punctuation, considering that all the words which you discussed are still present?

  2. iverlarsen says:

    Mike has answered the question generally, but was also asking for further examples.
    So, let me take up the other example that Jonathan mentioned: Luke 23:43.
    A word for word rendering without punctuation could be: And he said to him truly to you I say today with me you shall be in Paradise.
    It is usually not too difficult to deduce from the word order, normal patterns in Greek and context how such a sentence is to be divided. First we have a speech introduction: And he said to him. The beginning of the speech is “Truly to you I say.” (KJV is different because it is based on a different manuscript tradition, but that is not the issue here.) There is emphasis on “to you” since this applies to only one of the criminals, not the one who hang on the other side. Following this is what Jesus said to him: “Today you shall be with me in Paradise.” Since written Greek at that time did not use commas, colons or quotation marks, it has been suggested that “today” could be construed with the previous sentence and produce: “Truly I say to you today.” This is unlikely to be correct for a number of reasons. First, the expression “Truly I say to you” is a present tense. It is superfluous to add “today” and nowhere else in the NT is the common phrase “I say to you” combined with “today”. It would be unprecedented and not very logical. Furthermore, the fronting of “today” in the sentence “Today you shall be with me…” is explained by the relative emphasis on “today”. Grammatically, “today” can be placed anywhere in a Greek sentence: first, in the middle or last. The actual placement depends on the relative emphasis of the word. The more to the left it occurs the greater relative prominence it has. The criminal had said to Jesus: “Please remember me when you come into your kingdom”. He was thinking of a kingdom some time in the future. In contrast, Jesus responds: (Not some time in the future, but) Today…” This explains why the word is first in its sentence. The word “today” is also placed first in its sentence in Heb 3:15 and James 4:13. Some denominations and traditions prefer to include “today” in the previous sentence, but that is motivated more by theology than linguistics or the Greek text in context.

    Concerning quotations, it is normal in Greek to indicate the beginning of a quote by a speech introducer, a word for saying. However, Greek does not directly indicate the end of a quote. That is a matter of context. Usually there is enough of a break in the flow of the story to make it clear where the quote ends, but there are at least three places that come to mind, where the end of a quotation is disputed. One is John 3:16. Is this still part of Jesus’ speech or is it the beinning of John’s comment? Now, in the last decades we have learned more about how Greek functions in the area called discourse linguistics. Based on these fairly new insights and also a comparison of normal patterns in John’a gospel, it can be established beyond reasonable doubt that John 3:16 is not part of Jesus’ speech. NIV made a mistake here, but it has been corrected in the most recent edition. The same applies to 3:31 which is again the beginning of a comment by the author and not part of the speech of John the Baptist. Again the recent NIV has corrected the old NIV. The third example is the speech of Paul which begins in the middle of Gal 2:14. The RSV, ESV, GNB, GW and NET wrongly end the quote at the end of v. 14, while NLT ends it after v. 16. It should go all the way to the end of v. 21 as it does in NASB, NIV and some others.

    Another problem is that it is not always clear whether a sentence is a question or a statement. I’ll give the example of Matt 8:7. Most translations have decided that it is a statement: “I will come and heal him.” However, that does not match very well with the strong emphasis on I in the Greek text. It is possible and fits the context better to suggest that this was a stalling question: “Shall I come and heal him?” The assumption behind such a question would be that Jesus is responding to a Roman officer, a non-Jew, and Jews do not go into the houses of non-Jews. This scenario would be in the mind of any Jew, and Matthew was written for a Jewish audience. The question might be expanded: “Are you telling me that you want me to come to your house?” In this case, if I were to translate it as a question, I would need to put the traditional rendering in a footnote with an explanation of the possibiliites and reasoning.

    Finally, in adidtion to no punctuation, THEREWERENOSPACESBETWEENWORDS, and they used all caps. If you know the language you can almost always decide where a word break should be. One interesting ambiguity is in Rom 7:14. It could be either OIDAMEN, which means “we know” or OIDA MEN, which means “While I know”. Both could fit the context.

  3. Joel H. says:

    Punctuation is such an interesting issue, because modern text demands it even though it hadn’t been invented when the Bible was first written down.

    Three more examples from Hebrew may be helpful.

    1. Most importantly, the Hebrew leimor (“saying,” in common translations, but literally “to say”) introduces direct quotation, exactly like our modern quotation marks. So “God spoke unto Moses saying,…” should, in my opinion, be “God said to Moses, `…'”

    Both the Hebrew leimor construction and the English quotation-mark construction are the normal way of indicating direct speech, while “God spoke unto Moses saying” is barely English. (Compare, “my editor spoke unto me saying, when will I have your book proposal on my desk” — just to pick an example purely at random.)

    Furthermore, leimor is used not only for what is said, but also what is asked, sung, etc. So in Exodus 15:1, we find (ESV), “Then Moses and the people of Israel sang this song to the LORD, saying [leimor],…” But in English we don’t “say” songs. We sing them.

    2. The Hebrew halo introduces an exclamation, just like our exclamation mark at the end of sentence. And with exclamation marks, we can get a good sense of what’s going on. Spanish punctuation rules demand an exclamation mark at the end of the sentence, and matching upside-down one at the beginning. But, obviously, when we translation from Spanish to English, we don’t mimic the upside-down punctuation. And going from English to Spanish we add the extra punctuation.

    3. The Hebrew ha’im introduces a question, like our question mark at the end of a sentence.

    Finally, I would note that Hebrew didn’t have capital letters, either. But no one (that I know of) advocates leaving all of the English in lower case.

    I think that it’s just lack of knowledge that propagates the mistake of translating punctuation-words in Hebrew not as punctuation in English but rather as words.


  4. Theophrastus says:

    I’d like to note an implicit assumption in Joel’s comments — namely that the Pentateuch (and the entire Hebrew Bible) is primarily a written document (as opposed to an oral/aural document). The inclusion of extensive pointing (both vowelization and cantillation) in Masoretic Hebrew suggests that the Bible is primarily an oral/aural document. Further evidence is given by the ancient tradition of public Torah (and later Haftarah) readings.

    A translation that hopes to capture the oral/aural nature of the Bible will retain the marker-words included in Hebrew to the best of the extent compatible with English. Examples such as the KJV, Alter, and Fox show that this is possible.

    However, Joel’s remarks may be relevant when translating Mishnaic Hebrew or Medieval Hebrew. (Of course, Modern Hebrew has western-style punctuation.)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s