Skip to content
Tutorials Center
My learning

Topics

Lesson 67 of 86 ยท Deductive and Analytical Reasoning

Artificial Language, Logical Games and Truth Verification

Decode invented-language questions by comparing sentences, solve take-away strategy games with winning positions and verify self-referring statements by case checking.

Some reasoning questions feel unlike anything you have studied. A strange sentence in an invented language, a game in which two players take matchsticks, or a set of statements that talk about each other. They all test the same habit: find structure by comparing cases and ruling things out.

This lesson covers three of these formats. For each you get a method, worked examples and a small script that checks the answer. By the end you will know how to decode an artificial language from just a few sentences, how to find winning positions in a simple game, and how to verify statements that describe their own truth.

Part 1: Artificial language

In an artificial-language question you are given a few sentences in an invented language and their meanings. You must work out the word for each idea and translate a new sentence.

The method: compare sentences that differ in one place. If two sentences have the same words except one, that one word must carry the one difference in meaning.

Suppose we are told:

Invented sentence Meaning
tal vem sor big red house
tal nik sor big blue house
bix vem sor small red house
bix nik dun small blue car

Compare row 1 and row 2. Only the middle word changes (vem to nik) and only red changes to blue. So vem = red, nik = blue. Compare row 1 and row 3: tal to bix, big to small. So tal = big, bix = small. Row 1 and row 3 share sor, which must be house. Then dun in row 4 is car.

New sentence: "tal nik dun" is big blue car.

Two more habits help:

  • Word order may differ from English. If the invented sentences put the describing word after the noun, keep that rule when you translate.
  • Common words that appear in every sentence with the same meaning ("house" in each house sentence) are quickly identified.
JS Decode by comparison

Press Run (or Ctrl + Enter) to see the console output here.

No output.

The comparison step finds only the words that differ on their own. The elimination step then finishes the job: in row 1, once tal = big and vem = red are known, the only meaning left for sor is house; in row 4, once bix = small and nik = blue are known, dun must be car. Human solving mixes both tools.

Part 2: Logical games and winning positions

A take-away game has a pile of objects. Two players take turns removing some objects; the player who takes the last one wins.

Example. There are 21 matchsticks. Each turn a player removes 1, 2 or 3 sticks. Whoever takes the last stick wins. You move first. What should you do?

Work backwards. If 1, 2 or 3 sticks remain on your turn, you take them all and win. If 4 remain, you are in trouble: whatever you take (1, 2 or 3), the opponent can take the rest and win. So 4 is a losing position for the player to move. From 5, 6 or 7 you can take enough to leave exactly 4, putting the opponent in the losing position. Then 8 is losing again, because every move leaves 5, 6 or 7.

The losing positions are therefore the multiples of 4. From 21 you take 1, leaving 20. Whatever the opponent takes (a), you take 4 โˆ’ a, and the pile drops by 4 each round until you take the last stick.

flowchart LR
    A["21: you take 1"] --> B["20 left: losing for opponent"]
    B --> C["Opponent takes k"]
    C --> D["You take 4 - k"]
    D --> E["Pile is 16, 12, 8, 4, 0"]

Always leave a multiple of 4.

If the limit were 1 to 4 sticks, the losing positions would be multiples of 5. In general, with a limit of m per turn, leave a multiple of m + 1.

JS Find the losing positions

Press Run (or Ctrl + Enter) to see the console output here.

No output.

For a game where the player who takes the last item loses, the strategy shifts by one: leave a multiple of m + 1 plus one. Always test small cases first.

Part 3: Truth verification

Some questions list statements that talk about how many of the statements are true or false.

Example. Three statements:

  • S1: Exactly one of these three statements is false.
  • S2: Exactly two of these three statements are false.
  • S3: Exactly three of these three statements are false.

Which are true? Try each possible number of false statements.

  • If 1 statement is false, S1 would be true, which leaves two false statements: a contradiction.
  • If 2 are false, S2 is true and S1 and S3 are false: consistent.
  • If 3 are false, S3 would be true: a contradiction.

So only S2 is true. The method is to assume a count, check which statements that count would make true, and see whether the numbers agree. For liar and truth-teller puzzles about people, see Logical Connectives and Truth Statements.

JS Verify the statements by brute force

Press Run (or Ctrl + Enter) to see the console output here.

No output.

Tips

  • In artificial languages, look first for pairs of sentences with a single difference.
  • In games, write out the smallest positions first and work upward.
  • In truth puzzles, assume a count or an assignment and test it for contradictions.
  • Always read the winning rule: last takes wins versus last takes loses.

When stuck, shrink the problem. Solve the game with 5 sticks, or decode with two sentences, and the pattern usually shows itself.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. If 'pa ru' means 'red ball' and 'pa zi' means 'red cup', what does 'ru' mean?

    'pa' is shared with the same meaning, red. The differing words ru and zi mean ball and cup.
  2. 2. A pile has 25 matches; each player takes 1 to 3; the player taking the last match wins. To win going first you should take:

    25 mod 4 = 1. Take 1, leaving 24, a multiple of 4.
  3. 3. In a take 1-4 game where last takes wins, the losing positions for the player to move are multiples of:

    With a limit of m per turn, losing positions are multiples of m + 1, here 5.
  4. 4. Of 4 statements, S1 says 'exactly one of these four is true', S2 says 'exactly two are true', S3 says 'exactly three are true', S4 says 'none is true'. Which is the only consistent assumption?

    If exactly one is true and it is S1, then S2, S3 and S4 are false, which agrees. The others fail: if S2 alone were true, one statement would be true, not two; S4 true would contradict itself.

Summary

  • Decode artificial languages by comparing sentences that differ in one position, then use elimination.
  • In take-away games, find the losing positions from the bottom up; with a limit of m per turn they are multiples of m + 1.
  • For self-referring statements, assume a count of true statements and check for contradictions.

Next: Linear Seating Arrangements.