Roman Urdu
Urdu written in the Latin alphabet, with no dictionary to settle the arguments - and the reason our daily puzzles are built the way they are.
Urdu, written in an alphabet it was never built for
Roman Urdu is Urdu written with Latin letters. Not a dialect, not slang, not a simplified version of anything - the same language wearing a borrowed alphabet. For millions of people it is the only way they ever write their own mother tongue.
It grew wherever it was needed: phone keyboards that offered no Nastaliq, class notes taken in a hurry, chat windows, and the running commentary that follows every family decision. Nobody designed it and nobody approved it. It arrived by rough consensus, one message at a time.
Urdu is hard. Roman Urdu is harder.
People assume the difficulty runs the other way. Urdu looks intimidating - a script that flows right to left, letters that change shape depending on their neighbours - so surely familiar letters make it easier.
They do, right up until you need to be exact. Urdu has a settled script, settled spelling, and dictionaries that will tell you which form is correct: hard, but it has answers. Roman Urdu has none of that. Everyone who writes it solves the same problems privately and arrives at a slightly different answer.
So two people who pronounce a word identically can write it in two shapes and both be right. Charming in a group chat. A genuine design problem when you have to accept or reject somebody’s answer.
There is no correct version of a romanized word
We should say this plainly, because it shapes everything else here: we do not believe there is one correct way to romanize an Urdu word, and we are not trying to invent one.
The Latin alphabet simply lacks letters for what Urdu needs. Long and short vowels have to share five characters. Aspirated consonants - the ones spoken with a puff of breath - need a mark English never developed. The retroflex sounds, made with the tongue curled back, have no letter at all, and neither does the nasal ending that closes a word (Noon Gunna - we are staring directly at you).
So every writer improvises, and none of those choices is a mistake. They are reasonable people solving an unsolvable problem in the two seconds they had.
One idea, five ways to say it: “your father”
Here is the example we return to whenever someone says this should be straightforward. Take two ordinary English words, your father, and say them the way you actually would at home.
You might say aap ke walid, which is what you write on a form. You might say aap ke abbu, warm and respectful and by far the most common thing you will hear. You might stretch it to aap ke abba jaan, which carries more affection and usually more age. You might drop to tumhare baba with a cousin. And you might say tera baap, which between two people who grew up together is pure warmth - and between two strangers is the opening line of an argument.
Two words in English, and five different expressions in Urdu - not five spellings of one phrase, but five separate ways to name the same person, each carrying its own distance. Spelling is the argument that starts afterwards.
And “your” has to answer two questions at once
English gets one word for this. Urdu makes you settle two things before you can finish it, and they are independent. The first is who you are talking to. Aap ka is respectful, tumhara is for a friend, and tera is for someone you have known since you were both small - or for someone you have decided to be rude to, which is honestly also a valid use of it.
The second is what you are talking about, because the ending agrees with the thing being possessed, not with the person you are addressing. It is aap ka beta and aap ki beti no matter who is standing in front of you, and the same swap gives tumhara or tumhari, tera or teri.
The third form, ke, is the plural, and respect borrows it: speak about someone deferentially and the grammar treats them as more than one, so the ending follows the noun into the plural. That is why the examples above read aap ke abbu rather than aap ka abbu. Nothing about aap causes it - the same shift produced tumhare baba a few lines earlier, and it gives tere in the most familiar paradigm of all.
The first ladder is invisible in English and unmissable in Urdu: one rung too far down and a polite sentence becomes an insult, and a learner will not find out until somebody’s face changes. The second is pure grammar, and it catches people who already speak the language.
One word, five spellings: a short field guide
Take the most ordinary word in the language - no. You will meet it written nahi, nahin, nahee, naheen, and, from anyone typing at speed, nhi. Five forms, read aloud identically, and not one of them wrong.
Every difference between those five comes from a decision below. If you and a friend spell the same word differently, it is almost always one of these, made independently and years apart:
- A long vowel written double or single, because the alphabet offers one letter for both lengths
- An h that marks a puff of breath rather than a sound of its own
- The same ending written with ee or with i, or the same rounded sound written with oo or with u
- A k where somebody else writes a q, for a distinction many speakers no longer make aloud
- A closing n that is nasal rather than pronounced, so half of us write it and half do not
- A letter kept because the Urdu spelling has it, or dropped because nobody pronounces it
How we handle all this inside the games
Our answer is not to declare a house standard and make everyone learn it. It is to collect the variants during review and accept as many as each game structurally can.
Some games have room to be generous. A guessing game can ignore ordinary capitalization and punctuation and accept several spellings, because nothing else on the board depends on which one you chose.
Others cannot. In a crossword the spelling is load-bearing: the answer has a length and every letter is checked in both directions. So the grid decides, and the crossings tell you which spelling this puzzle wants. That is not us insisting a spelling is correct - it is geometry.
A person reads every puzzle before it goes out
Technology helps us research and draft, but nothing reaches you without a human reading it first - the clue, the answer, the difficulty, and whether the whole thing is actually fair.
Most of that time goes to spelling. Writing a clue is quick; listing every spelling a reasonable person might type, then deciding which the game can accept, is the slow part - unglamorous, and the reason a puzzle feels fair rather than fussy.
It also means our mistakes are human ones. We miss variants, and we favour the spelling we grew up with without noticing.
Thank you - and please tell us when we are wrong
Guessistan has grown a great deal in a short time, and it happened because people played, shared their results, and argued with us about the answers. Every day somebody spends a few minutes here instead of anywhere else online, and we are genuinely grateful for it.
So we want the whole range of feedback, not the polite half. Tell us it was too easy, the clue was unfair, or the spelling we accepted is not the one anybody in your family uses. Send the regional variant we missed and the game you wish we would build next. Criticism is more useful than praise, and far cheaper than learning it from a player who simply stopped coming back.