# Can magician apprentice know what his master is writing just by hearing it?

We have a magician apprentice (let's call him Steven) and his master (Destionell the Great). Steven is serving for a long time to his master and he truly hates it.

Every day all his job is just cleaning Destionells house, cooking him meals and washing his robes. And all of that without the magic of course (because you're not ready yet, the master always say to Steven).

The apprentice actually never learnt a single spell and never saw a single book of them (the books are even enchanted so no one could open them without magicians knowledge), but Steven desires a hopes to be a great wizard, so he's not leaving his master despite all of this. Steven never saw his master casting any spell, but he believes him strongly, as he is well-known magician.

One day Steven came up with a plan. Because he was often delivering masters mail to the nearest village, he has access to some (unimportant) handwritings of his master. Also, he is sometimes cleaning the room when Destionell is writing some spells (with a regular pen on a regular paper like the magicians always do).

So Steven decides to learn how all the pen strokes and the gasps between words his master does sound. He will always listen very carefully and after a certain time, as he hopes, he will be able to find out more about some of the spells his master is writing, just by hearing.

TL;DR: Is it possible to learn to recognise what someone is writing just by hearing it (if you are practising with the same person)?

• In that case, no. Listen to someone typing on a keyboard. Can you tell what they're typing? Maybe the spaces and enters, but not the regular key strokes. Same with writing. There's no reasonable way for you to tell what is being written purely by listening. Steve needs to get sneaky and find a way to look over his master's shoulder for that one. Jan 5 '17 at 17:20
• @AndreiROM That's not entirely accurate. Each key stroke does have a different sound, but they're effectively indistinguishable to the human ear. Jan 5 '17 at 17:21
• @TGar listening to writing on paper would never work. Do you realise how often people stop writing in the middle of a word because they're thinking? I know that I do that very often, or even in the middle of sentences while typing. Your apprentice would be very very confused by the pauses.
– Aify
Jan 5 '17 at 17:29
• @TGar for example, consider English vs Chinese. The complexity of Chinese characters would probably make it much harder to pick up from hearing it written. Jan 5 '17 at 17:36
• Sorry I don't have an answer to your question. My comment was on behalf of other WBers. The world is on a 24 hour cycle. It's just that it's always possible for better answers (from others) might turn up. Why answer a question when an acceptance has been lodged? Patience can be a virtue. Jan 6 '17 at 9:21

# Absolutely not

If the master is writing for himself, he's probably using very casual handwriting, the sort that would be used when writing letters to friends. This handwriting is going to be in cursive - and cursive is written in a continuous stroke. If Steven had a supercomputer, a massive archive of perfectly reliable data, and the ability to write machine learning software, he could maybe analyze the sounds of individual words, but he will never get letter accuracy.

And then you get to the other problems inherent in writing on paper. The magician crossed something out or underlined something. Where? You have no idea. The magician paused in the middle of a word. Is it a new word when he starts writing again? You can't know. The magician scribed a diagram? Good luck reproducing it.

The master wrote something in bigger size? The sound just changed. The master is writing on different paper, or with a freshly sharpened quill? The sound just changed.

Add to that the fact that Steven can't sit there writing it down as he hears it. He needs to have 100% perfect memory, and even then he's guaranteed to screw up and accidentally open a portal to the Far Realms or something.

• Because we all know how reliable the brain is when it comes to memorizing sensory impressions. Or wait, it's the opposite of that. Jan 5 '17 at 18:05
• You don't need reliable memory to recognize language, because it's full of redundancy. Jan 5 '17 at 18:13
• Even deciding that the master using a specific single language relies on a huge swath of assumptions that Steven can't afford to make. Jan 5 '17 at 18:15
• Steven can also assume that the master is writing the word "butts" over and over, in which case the answer to the question is "yes, easily." Jan 5 '17 at 18:20
• You assume all casual letters are written in cursive. I think you're too quick to dismiss the idea. Aug 25 '20 at 20:48

It should be possible, even though extremely hard.

It is a somewhat-common party trick to recognize what other person is writing with wet finger on a glass table without looking at it, but the sound produced is quite loud, and the other person is purposefully writing slowly.

Recognizing pen writing should be possible using the same principle, although much harder, as sounds are more quiet, and magician is probably typing faster, which would make it harder. Than again, paper grain should make vertical strokes easier to tell from horizontal ones, if your ear is keen enough to hear them at all.

Some magical hearing-enhancement drug should make it a breeze, but it is not "impossible" even without it.

Limit here is that apprentice would need to know language magician is using - letter recognition would not be perfect, and he would often need to rely on context to tell whole words, which makes magical incantation in ancient languages impossible to decipher with 100% accuracy (plot device detected).

# Yes

It is certainly possible for Steven to figure out what his master is writing, with a few foundational assumptions.

1. Steven must have a very acute sense of hearing, a musical background will be of great assistance.

2. The language and general writing style used in the casual letters and notes that the Wizard uses are the same, otherwise it would be impossible to learn how to compare and correlate the stroke sounds to the writing.

3. Steven is either given enough time to perfect this skill to the point of being able to speak out the words just based on the strokes, and/or has Eidetic/Photographic -grade memory so he can reconstruct passages later (especially true for alphabet-based writing). Alternatively, he could just write as he listens from a different room.

4. The writing is done in a language with some sort of pen-stroke system, and strokes are done in a sequential order to create letters/words. More importantly, each word/letter should have mostly distinct strokes. (Cursive is a special case that does not fall under this category, and surely if his Master writes in cursive, Steven isn't quite mad enough to attempt to distinguish between entire words in cursive by sound alone).

5. Credits to AlexP for mentioning that some languages, like Arabic, are strictly done in cursive form. If Steven's master is writing in an Arabic-like or strictly cursive language, Steven should just give up on this venture and find a master that writes in a different language.

6. There is a distinct enough gap between letters when writing, otherwise there will be a lot more work involved in figuring out which strokes belong to which letters.

# Process

## Stroke Categorization

Steven must begin the process by figuring out a system to categorize the strokes.

Example Alphabet:

As can be seen from the image, each letter of the alphabet has a distinct number of strokes, and in a set order (which some individuals do differently, but the order for each letter can be analyzed through seeing and listening to the pen strokes of the Master’s casual letters).

The strokes forming the letters of the alphabet can be categorized into a few distinct types: straight, curved, dotted, and angled strokes. The length of each stroke can be determined by listening to the duration of each stroke: short, medium, or long.

If Steven can train his ears to hear the differences between each type of stroke and the length, using this system he can construct a sequence of writing sounds into a series of strokes, letters, and subsequently sentences.

## Example:

S- straight stroke
C- curved stroke
D- dot stroke
A- angled stroke

1- short
2- medium
3- long

"A quick brown fox"

s3s3s1
A

c2s3 c1s1 ds1 c2 l3a3
q/g  u/n  i   c  k

l3c2 l1c1 c3   a2a2 s1c2
b/d  r    o/s  w    n

s1c3 c3   s1s1
f    o/s  x

Autocorrecting process:
1. Agnickdrownfsx
2. A quick drown fsx
3. A quick brown fox


## Practice and Results

By piecing together each letter combo, and eliminating the incorrect letters, Steven can reasonably hope to piece together entire sentences, of course, with some errors.

It will take a long time before he could do it for everything that is written, but when he masters this, it should be enough for him to begin his journey into magic, so that he may catch the attention of his master and finally get a chance to learn in earnest (or at least animate some brooms to do his chores).

# Bonus:

## Symbol-based language:

It may seem more difficult to deal with a symbol based language than an alphabetical one, but in actuality it can be made much simpler, since each stroke type in these languages are more well-defined (see how each stroke has various defining features).

Assuming, every 'word' constructed of such strokes is always written in the same order, it is possible to construct it in unique, less repetitive strokes compared to a phonetic language, with much less mistakes in determining which word/symbol has been used. The words can then be recombined to form sentences.

• That Latin alphabet on top is not how most people who write longhand write longhand. Normally, people who actually write longhand write in continuous strokes forming most of a word, with details such as bars across tees and dots over eyes being added afterwards. Here is an example. Aug 25 '20 at 21:00
• In that case it will be harder, with the timing between words being more important. Aug 25 '20 at 21:03
• The example you have given is practically cursive though, I do not believe people who want to have legible notes will opt to always write in such a fashion. Aug 25 '20 at 21:03
• I have highlighted one of the conditions of being able to listen in as distinct sequential strokes, as otherwise there would be no way to distinguish anything. Luckily even in almost cursive longhand, as long as the person writing follows a set routine when writing, and has enough gaps between letters, it'll still be possible to distinguish between groups of letters, to some extent. It'll just be a lot more difficult to guess at what is being written. Aug 25 '20 at 21:08
• ? The example is perfectly legible -- a regular and clear hand, as they say. Most people in most European countries learn to write like that; there are of course cultural variations (for example, Romance countries use a different shape for "p" and "r"), but by and large we all learn this script. In Russia they learn a sort-of similar script with Cyrillic letters. And most importantly this is how people wrote longhand for many centuries; this script became common roughly at the beginning of the Renaissance. Aug 25 '20 at 21:13