How to Solve Linguistics Puzzles With Unfamiliar Data

Learning how to solve linguistics puzzles starts with getting comfortable making a small observation before you have an answer. If none of the words look familiar, you can still tell whether two strings end the same way or whether two translations differ in just one detail. Those are clues you can check without knowing the language. The aim of your first pass is not to explain the whole page. It is to find one claim you can test.
How to solve linguistics puzzles one repeat at a time
Start with the examples the puzzle has already matched to translations. Mark an exact repeat in the unfamiliar forms, then look across at the corresponding translations. Do they share an object, a number, a location, or some other detail? If you see a repeat but cannot find a shared detail yet, leave the mark and move on. You do not have to explain every repeat you notice.
Use this made-up dataset to try it. For now, read only rows 1 to 4. Cover the last row with a sheet of paper or scroll so that it is out of view. You will use it to check a prediction, so it helps not to see the answer yet.
| Row | Form | English translation |
|---|---|---|
| 1 | paku | one basket |
| 2 | pakumi | two baskets |
| 3 | selo | one cup |
| 4 | selomi | two cups |
| 5 | navi | three cups |
You can mark paku in rows 1 and 2, selo in rows 3 and 4, and the ending mi in rows 2 and 4. Write those repeats down before you assign meanings to them. The note “mi appears at the end of rows 2 and 4” is an observation. “mi means two” is a guess. Keeping those statements separate makes it easier to revise the guess without throwing away what you actually saw.
Choose two examples that change in one way
Compare rows 1 and 2. The object stays the same: both translations are about baskets. The number changes from one to two. In the unfamiliar forms, the only visible change is that row 2 adds mi at the end. Rows 3 and 4 give you the same comparison with cups. English also changes basket to baskets and cup to cups, but that does not add a second change in meaning. It is another way English shows the change in number.
Now try rows 2 and 3. Both the object and the number change, so it is harder to tell which difference in the forms goes with which difference in the translations. That pair is not useless. It just asks you to sort out two things at once. When you have a choice, work with rows 1 and 2 first.
If the puzzle gives you no pair that differs in just one detail, look for two comparisons that share a change. For example, if two comparisons both change the number but each changes a different extra detail, check whether the same piece appears in both changed forms. Treat the result as a lead rather than as proof. On your scratch paper, keep a short record of which rows you compared and what stayed the same.
Make a prediction before checking the unused row
From the first four rows, you might guess that mi marks exactly two. You might also guess that it marks any number greater than one. Both guesses fit the rows you have read. Checking either guess against row 4 again will not help you choose between them. You need a row on which they predict different things.
Try making a prediction for a translation with three cups. If mi marks any number greater than one, you would expect the form for three cups to end in mi. If mi marks exactly two, you would not expect that ending. You do not need to predict the whole form. A prediction about the ending is enough to check the difference between these guesses.
Now uncover row 5. The form is navi, with no mi ending. That rules out the guess that mi appears with every number greater than one in this dataset. It does not prove that mi marks exactly two in all cases. Row 5 also lacks the repeated form selo, so the data do not establish how to build the form for three baskets. A useful note at this point is “mi appears in both examples with two, but not in the example with three.”
Keep your first rule small enough to revise
When a prediction fails, resist the urge to label the row an exception and move on. Write down exactly what failed. In this example, the failed prediction was about the ending, not about whether the translation described cups. If you revise only the part of your guess that the evidence challenges, you can keep working with the observations that still hold.
Use the ILC training and practice problems to try this on a fresh dataset, or browse more exam preparation articles for other ways to practice. On your next puzzle, leave one matched example unread if the layout lets you. Beside your first guess, write a prediction specific enough to mark right or wrong, such as “The form in this row will end in mi.” If the unused example fits both of your guesses, look for a different row rather than counting that match as a decision.
Frequently asked questions
What if the same letters show up everywhere?
Check where they appear and which translations go with them. A repeat is more useful when its presence or absence lines up with a specific change.
What if I already read every example?
Use a row you did not rely on when forming your guess. Write down what your guess predicts for that row before inspecting it again.
What if both of my guesses fit every row?
Keep both for now and write down what each would predict in a case where they disagree. If the puzzle provides no such case, do not treat one guess as proven.


