A Three-Mark Translation Where the Clause Order Reverses the Meaning
A Std X Third-Language Gujarati teacher is marking the 3-mark translation question: 'Translate into Gujarati: We reached the station after the train had left.' The rubric is meaning 2, language 1. Of 48 scripts, 12 wrote 'ટ્રેન ગયા પછી અમે સ્ટેશન પહોંચ્યા.' Twenty-one wrote 'અમે સ્ટેશન પહોંચ્યા પછી ટ્રેન ગઈ.' Nine wrote 'અમે સ્ટેશન પર પહોંચ્યા ત્યારે ગાડી જતી રહી હતી.' Six wrote a hybrid, ending with a Devanagari form: 'અમે સ્ટેશન પહોંચ્યા પછી ટ્રેન ગઇ हती.' Several students have parents who speak Marathi at home, and the teacher knows that clause order differs between the two languages. The scripts must be returned with marks the day after tomorrow, and the same class sits a translation question again in the terminal exam.
1. Mark the 21 reversed-order scripts. What exactly has gone wrong, and how do you satisfy yourself before awarding zero for meaning?
Read the worked answer
The error is temporal, not lexical. 'અમે સ્ટેશન પહોંચ્યા પછી ટ્રેન ગઈ' says the train left after we reached, i.e. we caught it or nearly did. The English says the opposite. Every word is correctly translated and the sentence still means something different - which is precisely why translation is tested with more than a vocabulary check.
Before awarding zero on meaning, apply the back-translation test: read the student's Gujarati sentence aloud and put it back into English. If it yields 'The train left after we reached the station', the meaning is inverted and the 2 meaning marks cannot be given.
What of the language mark? The sentence is grammatical, so 1 mark for language is defensible, giving a total of 1. Some markers would give 0 on the ground that the sentence is not a translation at all; pick one rule and apply it to all 21, and write the rule on the scheme. My choice is 1, because the item is scored on two separate criteria and one of them is genuinely met; the mark profile then tells the teacher that the student can write Gujarati but is not reading the English clause structure.
2. Decide on the nine scripts using 'ગાડી જતી રહી હતી' and the six with the Devanagari ending. Justify both decisions.
Read the worked answer
The nine 'ગાડી' scripts deserve full marks. ગાડી is ordinary Gujarati for a train, alongside રેલગાડી; refusing it would be penalising a correct register choice, not an error. Their structure is also right: 'પહોંચ્યા ત્યારે ... જતી રહી હતી' preserves the sequence, with the departure already complete at the moment of arrival, which is exactly what the past perfect conveys. 'જતી રહી હતી' is an acceptable rendering of 'had left'. Award 2 for meaning and 1 for language where the rest is clean.
The six hybrid scripts are a different matter. A stray Devanagari 'हती' shows the candidate has not yet automated the Gujarati script; it is a script slip, not a vocabulary error. Charge it to the language mark, not the meaning mark: meaning 2, language 0, total 2. If the Devanagari form made the word unreadable, or if the sentence were half in Devanagari, then comprehension is in doubt and the meaning marks would be affected. Draw that line explicitly on the scheme - scattered single letters cost the language mark; sustained mixing costs meaning - so that all 48 scripts are judged by the same rule.
3. What should change in the teaching and in the item before the terminal exam?
Read the worked answer
Teaching: the class needs contrastive work on clause order, not more vocabulary lists. Give six English sentences with after, before, when and until, and have students first draw the timeline, then write the Gujarati. The timeline makes the reversal visible before it reaches the page. For the Marathi-speaking students, spend five minutes naming the difference explicitly: Marathi allows participle-led order that maps neatly onto English, and Gujarati prefers the પછી-clause first; students who translate structure rather than meaning are the ones who invert.
Item design: keep the 3-mark split (meaning 2, language 1), because it produced useful diagnostic information - the 21 scripts differ from the six in exactly which criterion failed. Add a stated policy on script and spelling to the rubric. And pre-test the sentence on two colleagues, one of whom should back-translate it; if their back-translation is ambiguous, the item is ambiguous. Finally, include one sentence in the terminal paper whose meaning survives reordering, so that the section is not solely a trap for clause order.