Assessing NT students on translations of NT texts is a waste of time

Asking New Testament students to provide a translation of a known (chapter/verse) NT text in an exam or exegetical paper is a waste of time. It tests nothing and it discriminates nothing.

Every student ought to be getting 90-100% on this part of an assessment anyway, because either:

  • they are smart enough to check any translation they do with several English versions and realise their errors beforehand
  • they are smart and a little unscrupulous and are just going to vary an existing English version anyway.
  • if it’s an exam situation, and it’s a set text, then all we are testing is their preparation, not their ability to read Greek.

Why are we even asking them to do translations anyway? They are unlikely to create a translation that is genuinely better or meaningfully different from the hyper-abundance of English versions already in existence. And, assuming that this is a paper and not an exam, as an examiner I gain almost zero insight into their Greek ability or their understanding of the text, unless they stuff it up horribly. Even on an exam, I am testing an unrealistic environment and their ability to read Greek under pressure but with the real question of “what did they prepare and how well?”

Consider what translation is meant to be testing.

  1. Their ability to read Greek.
  2. Their ability to translate
  3. Their understanding of grammar.

In reverse order:

Understanding grammar: Yes, it’s true that asking someone to translate something is a fast and relatively effective way to see if they understand it. But a set translation doesn’t do that, and a set text doesn’t do that either. Knowledge of the English is too strong an interference.

And, if we are asking students to translate in a way that demonstrates their knowledge of the grammar (as I unfortunately have to instruct some of my tutoring students to do), we are training them to be bad translators.

If you want to test explicit grammar knowledge, just ask explicit grammar questions. It’s that simple.

Ability to translate:

Translation is a skill, and it’s a high order one, not a low order one. It requires a real and comprehensive ability in both the source and target languages. Most translation exercises are a test in the students ability to reproduce “Biblish” English.

If we genuinely wanted to test (and train!) translation ability, we would set translation tasks/assessments in which the translation target was specified in demanding ways: Basic English, Anglish, a specific Flesch Reading Ease score, Poetry, highly idiomatic, highly regionalised, etc., etc..

Ability to read Greek:

In some ways it’s difficult to easily, accurately, and comprehensively assess someone’s ability to read Greek. It’s certainly possible to, for instance, ask questions in Greek. But then you are also testing their ability to comprehend the questions. And, you run into ‘form’ problems – if you understand the form of the question, and the answer, you can answer a foreign language question without actually understanding the question or its answer. I did this a great deal in Mongolian because of my “fake it till you make it” philosophy.

I think you might till test this via translation, if you want to, but by varying a few factors. Firstly, in exams simply remove all chapter and verse markings. maybe even punctuation. You don’t need to make them read undivided uncials though, unless you are cruel.

Secondly, you could also provide texts that are single-manuscript sources, rather than eclectic texts. Students easily get used to the sanitised versions of ‘text’ that critical editions give them. Throw them into the wild and have them read something more like this:


ειπ]ε[ν ουν] αυτοις ο ι̅η̅· α[μην αμην λεγ]ω̣ υμιν̣ [οτι εγω ειμι η θυρα των

π̣ροβατ̣ω̣ν· παντες οσοι̣ [ηλθον κλεπτα]ι εισιν και λησται· α̣[λλ ουκ η

κουσεν αυτων τα προβατα[· εγω ειμι η] θ̣υ[ρ]α· δι εμου εαν τις εισ[ελθη σωθη

σεται· και εισελευσεται· και εξ[ε]λ[ε]υσεται· και νομην ευρησει· [ο κλεπτης

ουκ ερχεται ει μη ϊνα κλεψη και θυση και απολεση· εγω η̣[λθον

ϊνα ζωην εχωσι· και περισσον εχωσιν·


(John 10:7-10 from P45)


If they were classical Greek students tackling a NT class, you can be meaner. Convert a NT text into Ionian, or similar mean tricks.

Anyway, my point is simple – asking students to produce a ‘standard’-esque translation of a Greek NT text is not a very useful assessment tool because it doesn’t test anything very useful, and doesn’t provide any real discrimination among students. So it’s neither providing any valuable feedback to me as a teacher, nor is it spreading the student field in any meaningful way, so giving them all 10/10 is the same as giving them 0/10 for it. Yes, by all means students should probably go on translating NT texts as part of their studies, it won’t harm them (much), but let’s give up assessing them with it.

12 responses

  1. I agree that summative assessment on seen translation is problematic (not just for NT — when I was doing seen translation in exams for classical Greek I knew I was relying heavily on my memory of the English). And your’e right: quite apart from WHAT is being tested, discrimination is key in any summative assessment.

    I still think there’s formative value in getting students to work with texts they already know. But I deliberate say “work with” as I think there are interesting alternatives to translation afforded by computers, for example getting them to do a word-level alignment of the Greek to even a provided English translation. I know Greg Crane requires his students to align their translations—forces them to account for everything in the Greek.

    Other “seen passage” tasks might include getting them to disambiguate or provide missing morphological tagging, getting them to treebank a passage, etc. At intermediate-advanced levels, they can actually contribute to digital philology this way!

    I think I’ve mentioned this before, but I also think doing a translation of a seen passage from English into Greek would be fascinating to see the results of.

    • I think formative work is a whole different kettle of fish!

      Lots of scope for different kinds of thing, I like your suggestions.

      > I also think doing a translation of a seen passage from English into Greek would be fascinating to see the results of.

      Is a fascinating exercise. And can be done in several variations too! Translate into the style of Herodotus, Paul, Xenophon, etc.

  2. Pingback: Assessing NT students on translations of NT texts is a waste of time | Rdr. Thomas Sandberg

  3. Or you could assess them via essay. Our assessment is an essay in which they provide their own translation of a short passage of Greek NT, and then have to comment on it. This includes who they are translating for (college professors? people whose first language is not English? Teenagers? etc), and also requires them to comment on textual critical factors, comparison with other English translations (and why theirs differs), and to explore any words or phrases in more details that have particular theological/exegetical significance. I expect them to use appropriate resources (BDAG, LSJ, critical commentaries, apparatus etc etc). They can choose the passage (with my agreement/veto).

    For context, I am working in a setting where many of our students are training for accredited ministry in churches in England.

    I think this gives a more realistic task for assessment, and is closer to what I hope they will be able to do well by the time they finish the module. I agree that exams translating NT Greek are of limited value.

  4. I agree this type of activity allows people to cheat through getting it done. If our goal is rigorous academic assessment of a persons skills this is a problem. However is that the goal here? Perhaps the goal is to get people to practice bible translation? For me, having this task assigned to me and having the pressure of having it assessed was a weekly motivator to get it done—and I benefited greatly from having this weekly motivation. I just went back and test read a Greek passage I translated over a year ago and I can completely read/access what is going on in the text. There were only a few words I could not remember, but knowing the English filled it in.

    This is hugely encouraging for me. Having Greek text that I am not fully drowning in trying to understand is great for helping me move on to start thinking about grammar, discourse, etc… My initial reaction to this matter: I say leave it as an assessment task. Those who cheat through it are hardly going to succeed long term in academia.

    • I don’t suspect many NT students cheat to get this done, but that’s neither here nor there, it’s not my main focus.

      Nor is having it as one form of task that students do to improve their Greek. Again, I generally think that’s not a bad idea.

      My main point is that it’s not useful as an assessment task – it doesn’t provide much means of discriminating between students, it under-represents the difficulty and complexity of translation as an enterprise in itself, and it’s much more difficult to align the goal of testing with the task of translation, more difficult than a surface appraisal of what a ‘translation test’ might suggest.

      • I agree with you from the perspective of someone responsible for teaching. However I’m looking at it from the perspective of someone wanting to learn. For example, I chose to enroll in LA010 because one of its assessment tasks was to learn all Greek words that occur in the more than 10 Times—The assessment task lined up with my personal goals. I know those weekly deadlines are enormously helpful for me to help manage and focus my time.

        That said it would be cool if we could come up with better methods, but the question is what are they? i.e. i remember when I first started biblical Greek, one of my fears was oral assessment (because I had done lots of practice reading but not much in the way of conversation practice), I remember getting near the end of semester 1 being a little shocked when it finally dawned on me that it’s a written exam, and thus the only thing that was tested was ability to memorizing tables, words, and points of grammar. 😀

        By the way, re your sample text, it demonstrates perhaps a gap in ACT’s curriculum, I don’t believe there is any point where a student has an opportunity to become familiar with all those abbreviatoins and other oddities. As someone with some level of interest in textual criticism (both NT and LXX/MT) I feel it’s a gap. I am looking to see if I can work that into my research project for my final year.

        • I think this is confusing goals/methods/assessments. Which prompts me to write a post about it, but before we get to that, let me say briefly that deciding assessments for how they contribute to learning is back to front – one ought to work out 1) what students are to learn, then 2) how to effectively teach that, then 3) how to assess whether they have learnt it.

          So, it’s not about teacher v student perspective here, it’s whether using translation as an assessment effectively assesses what it is supposed to assess.

          As for my sample test, and the ACT curriculum, well LA010 could cover that kind of material. It’s up to the affiliate colleges how they instantiate ACT units. Though there are limits, obviously. The working assumption (reverse engineering from my external perspective) is that students who complete 1 year of Greek will “get enough” to then complete NT Exegesis subjects in Greek, and thus “learn enough” by the end of a few of them. This I would characterise as naive.

      • “1) what students are to learn, then 2) how to effectively teach that, then 3) how to assess whether they have learnt it.”

        I agree, that is why I think, (from a students perspective), understanding the assessments in a subject are important. In my experience, when I ask what the assessment tasks are, I get the impression (perhaps incorrectly) that the teacher thinks I am trying to work out if the subject is “easy” or not. However, what I am really doing, is trying to work out what a student is expected to know coming out the other end.

        For example, in Ridley’s “John” subject, we were allocated greek translation tasks and weekly handouts that discussed points of greek grammar (From David Black’s very helpful book “It’s still greek to me”). The greek translation work was assessed, and the grammar handouts were not. Every week most people submitted a translation, and I guesstimate that maybe 50% of the students prepared answers to the grammar handout questions.

        So the subject description could say (for example) that the goal of the subject is for the student to become familiar with the Greek text of John, and to become familiar with intermediate greek grammar, but the reality is, when a has 3-4 assignments due in the course of a month, coupled with a heavy involvement with their church, we all know what goes out the window first.

        • Oh, and absolutely. As much as teachers (falsely) believe that their methods determine what students spend their own time on, the reality is that what is assessed, and how it is assessed, very strongly shapes what students learn. Test vocab, students study vocab. Test translation, students study translation.

          Which is why good course design needs to align all 3 components. Definitely a post on that coming up!