The Language of Memory

This is a virtue which grants a single supernatural ability - The Language of memory. Unlike other abilities, this cannot be improved by exposure or practice.

The language of memory lets you write down a single memory using the special symbols of the language. Anyone who has this virtue can "read" your writing and create their own exact copy of the memory.

A single memory about something would take up an entire page from a book. A more detailed memory (or a duration longer than momentary) would need multiple pages.

The memory can contain any of the 5 senses, each sense adds more detail. The memory can contain information about an ability or language, but these are more complex matters needing a lot of planning and writing. As it is usually hard to separate knowledge, a person would write their entire knowledge of a subject in large block. After reading the complete block, all of that knowledge is gained.
Incomplete blocks only give fragments of a memory and conflicting information.

If a person has prior knowledge about a subject before reading one of these books they will need to make concentration rolls after the final page has been read. These concentration rolls determine if there are any conflicts between the new memories and the old ones.

Level 1 of the ability lets you write a single memory onto a single page.
Level 2 lets you use more complicated symbolism so you can put two memories onto a page.
Each additional level doubles the amount you can write in the same space.

A person cannot read information above their ability level - if they have level 1, they cannot read memories written using level 2.
The advantage of higher levels is that memories can be written more swiftly.
Higher levels can only be taught by someone who knows them. The current highest known level is 9 which was created by the founder. There are a few sets of tractati which introduce each level of the language.

We were playing around with numbers about how many pages were needed for 1XP of an ability, we were eventually looking at around 20 pages of LM 1 for 1 XP of an ability or art.

Level 5 in an ability (75XP) would take 1500 pages of LM1.

If a person already has level 4 in an ability they would still need to read all 1500 pages and pass a concentration roll to absorb the extra information.

The Language of Memory was created so that knowledge could be transferred in its entirety. It was created to preserve knowledge incase of a cataclysmic disaster such as a second schism war.

The Language of Memory is useful with Mental Construct laboritories because objects can be described using LM and transferred to other magi for use in their lab.

A magus with LM 9 could write 256 memories onto a single page. 6 pages could allow them to write 75XP worth of knowledge (1500 memories). These 6 pages can only be read by a person with LM 9 or higher

We weren't sure if we should follow the normal rules for reading text / writing text.

When magi have LM 9 it allows them to read and write memories very efficiently, they can exchange and learn knowledge very quickly - but the knowledge is the same. They cannot gain XP from reading a book the other wrote. It would be like reading a book they wrote themselves.

LM can be saga changing, summae would lose some value but tractati would become more precious.

One of the benefits of LM is to pass memories of events - especially for witnesses to a Hermetic Crime. People involved in such cases could have their memories transcribed in LM so that even if they die their testimony will remain.

1 Like

You invented a colourful new Supernatural Ability, and are thankfully very cautious about its impact on teaching Abillities.

Ars Magica takes care to prevent teaching Abilities via memory magic.

Look up Houses of Hermes: Societates (HoH:S) p.68ff Memory to see how this is done. It prevents creation or transfer of memory containing Abilities (Inscribed Memories and Procedural Memories).
The Mysteries Revised Edition (TMRE) p.25ff The Art of Memory shows an approach to memorize entire books, extending real memorization techniques by Hermetic magic. But it takes still time - typically seasons - to properly read and understand these books that are now in your memory.

The latest Sub Rosa, #22, has an adventure based on a magus trying to push the envelope about teaching knowledges with magic.


on that note I will say that I don't see memory transfer being an issue for this form because it seems to me it would take longer to "read" the memory than to experience it first hand, where standard writing of books is a summarization of knowledge distilled from experience. At best I could see this being used to create a sort of realia from moments in adventures that contributed to understanding an ability...

Thanks for these references. Unfortunately my books are half-way around the world these days, so I can't look them up. I would like to point out that we aren't really teaching abilities by magic - we are writing them down in detail. The person on the other end has to read them, which can take many seasons of reading.

The example of learning a skill in only a few pages was when it was taken to the extreme - where both writer and learner have level 9 in the language.

When I get access to my books again I will be able to read the relevant chapters.

Hi Silveroak,
This ability can definitely be used to store "teachable moments", such as battles that were fought or strange applications of magic to solve problems.
Being able to pass on a full memory of what happened will be really useful when teaching apprentices. (or bragging)

Hi OneShot, Thanks for those page references, they were very helpful. My friend sent me a text extract from a PDF about the memories from Societates. From what I can see, my supernatural virtue would allow people to write down both Episodic and Inscribed memories for people with the virtue to read. The virtue allows people to naturally make their own "copy" of the memories.

The text notes that it should be possible by hermetic magic, but right now they can't do it. I think that this is in-line with most other supernatural virtues.

I don't want to break the game, but I do believe that this can add to the game if it is implemented well.

For LM to be practical in game it needs to be superior to reading summae and tractati, but it needs to have serious limitations too.

I think that what needs adjusting or balancing is the reading/writing part using the higher levels of the language of memory.

Perhaps each advanced symbol takes a longer amount of time to write as the magus has to choose the correct symbol to represent the aspect of their memory.
Perhaps the reader has to concentrate on each symbol as they read it to assimilate it properly into the right place in the memory (like fitting a jig-saw together - even with instructions it takes a long time to do it).

Does anyone else have an idea on how to balance these things?

Another approach is to alter the Language of Memory ability, so that it determines how fast you can write, nothing to do with the word/page count. We can switch it to use the standard text rules from Ars Magica. The bonus to this is that it allows the full level of the ability to be written, not half of it. The downside is that the text must be read in full before the knowledge is gained.

So, if we look at the example magus from the main book.
Quintus has a score of 24 in Ignem and -1 Communication and Latin 5. If he wrote a standard summae he could write a level 12 quality 5 summae in 3 seasons.
How can we change this using LM?

Well, for a start he would now be writing a lvl 24 summae - but what about the quality? The "normal" convention of Communication + 6? I'm not sure about this because it takes too long to read a text.

How long does it take him to write it? Do we use the normal convention, Language + Communication?
He's not writing it in Latin, so his Latin score isn't used, it would have to be his LM.
LM of 1 + Communication -1 = 0. So he would need an LM of at least 2 to achieve anything.
LM 2 would take him 24 seasons to write a book. (1 point per season)
LM 3 = 12 seasons (2 p/s)
LM 4 = 8 seasons (3 p/s)
LM 5 = 6 seasons (4 p/s)
LM 6 = 5 seasons (5 p/s)
LM 7 = 4 seasons (6 p/s)
LM 8 = 4 seasons (7 p/s)
LM 9 = 3 seasons (8 p/s)

How about the reader? Art lvl 24 is 300XP at a quality of 5 this gives 5XP per season that takes 60 seasons to read it. That is a long time to read a single text - especially when you don't get any benefit from it until the end.

XP shouldn't be gained from reading it. We need a similar process as for writing it. Once enough progress is made it has become read. So we need to generate completion points.
How about we use the same points for the author and the reader. Basically the length of time needed to write it is how long it takes to read it in full? Or do a similar calculation for the reader - LM+Communication, or even just LM.

So I can see this virtue developing in two different ways. What do you guys think?

I think the system you propose is not unbalanced.

I would certainly be wary of spending 60 seasons to read a book even if I get 300 xp all at once at the end of the 60 seasons.

On the other hand it could be better if the author had a higher communication. I personally dont have any problem with that. It is after all a new supernatural ability so it is fine if it improves on something and/or opens new possibilities. IMO that is the whole idea in inventing a new supernatural ability.

There is a another published example of something related, the Calc rune effects in Thrice Told Tales. Only a little bit of it is relevant bit it essentially lets a rune wizard put CrMe/Im style effects into a rune script, so that anyone who can read the runes has a memory or story imparted to them.

The higher level guidelines allow enough to be stored in the script that it acts as a tractatus or summa. The limitation is that these take as much time to "write" as a book would, and count towards the max number of tractatus a character can write. They also take a season to study just like a book.

Not an exact match, and not a hermetic effect either, but maybe of interest anyway.

Based on that a possible way to balance it is to do something similar to the progression of those guidelines, but with this supernatural ability. To make higher levels of the supernatural ability worthwhile maybe allow a bonus to the quality (but also limit who can read it). Something like (and this is just spitballing, to give an idea of what I mean):

LM 1 = Can write memories lasting a few moments.
LM 2 = Can write memories lasting a few minutes.
LM 3 = Can write memories lasting a few hours.
LM 4 = Can write memories equivalent to a tractatus (using his LM score as the language).
LM 5 = Can write memories equivalent to a summa (using his LM score as the language).
LM 6+ = Can choose to add a bonus of (LM score - 5) to tractatus/summa quality. Only a reader with a score in LM greater than or equal to the author can read the text.

So for example Quintus there with LM 9 writing his Language of Memory Ignem Summa would end up spending 2 seasons for a level 12 quality 8 summa which can only be read by people with LM 9 or higher. He could also spend another 5 seasons writing tractatus equivalent memories down (his tractatus limit at Ignem 24) and each would be quality 8 with the same limitation as above.

I think that communication should still play a factor in the final quality when studying memories for xp - you (the author) still need to be able to string memories together coherently in a way that is comprehensible and teaches the right things in the right order, which to me would fall under communication.

Maybe that's too far from the original idea though. It's certainly an interesting idea for a supernatural ability if you can get the balance right.

I would not think that any ability which allows a summae to be written at full ability would be balanced, since that gives the same upper limit as teaching in a single text, and then allows tractatus in addition. 2/3 might be a better goal.
However, it also occurs to me that again teaching by copying experiences would be less efficient than normal teaching, unless the memories being transcribed are of learning directly, in which case the lecture commentary from A&A might be a better model than writing summae- essentially a tractatus that does not count against the level restrictions total, and perhaps add the entire ability in the language of memory to the SQ instead of scribe/2, or perhaps retain half of the teacher's teaching ability.

Precisely, the text says:

Hermetic magic should, in theory, be able to duplicate procedural and inscribed memories from a donor to a target, allowing the transfer of Abilities. At the present time, no magus has accomplished this using conventional spellcraft.

In other words, it leaves it indeed to SG and troupe to decide, just why Hermetic magic currently cannot do it. Is magic theory just too optimistic and it actually can't be done? Or are magi currently just not skilled enough to work it out?

When thinking about balancing it, keep in mind that a character just created can rather easily have an Ability score of 9 in a Supernatural Ability, if she takes Puissant (ArM5 p.48) and Affinity (ArM5 p.40) with it. So what does a group of characters, all with a score of 9, do with your saga? Does that balance the xp and the points of Virtues?

1 Like

Actually they could have 7+2, which matters in this situation since for instructional purposes their ability is 7, not 9.

It is 7 for learning / teaching / writing about the Supernatural Ability (see Errata), not for learning / teaching something else with it.

You are particularly adept with one Ability, and add 2 to its value whenever you use it. Note that you do not, in general, use an Ability when learning it, teaching it, or writing about it.

An ability score of 9 in a supernatural ability does not imply that teaching is the ability. I would not expect a new supernatural ability to be posessed at level 9 by all characters unless it could be taught to level 9 by one of those characters, which it generally cannot.

I do notice however that with the lecture commentary rules a character can easily write more experience in texts in an ability than they themselves possess...

If at the start of the saga a Supernatural Ability at level 9 in a group beats the other ways to spend xp and points of Virtues, you of course need to expect such a group. Or - better of course - reduce the power of that Supernatural Ability.

One virtue to have the supernatural ability, one virtue for puissant, one virtue for affinity, I would expect one person to max the virtue then start training others in it, where the don't have to spend the virtues.

First, I want to thank everyone for their thoughts on this topic. I am grateful for all of the people who took time to comment. You are helping to point out the weak points in my idea and make it fit into the setting.

I agree that the author having a higher Com will make things faster for everyone, I just used the example from the core text.

Having a Com of +1 or +2 would change those numbers a lot. But even then it is still the same as reading a normal summae. Except you only get the benefit at the end. Although you can write/read a book to a much higher level than normal.

@Argentius Using those rules from the Calc Rune looks interesting, I like the idea of the different levels of LM. What sort of time frame are we considering for writing these ideas? Does the Calc section give any guidelines? I could see something in LM1 being written in an hour, LM2 a few days, LM 3 a month.
The tractatus part I'm not sure about - the point of LM was to write knowledge in its entirety. But I really like this direction.

@silveroak once you have finished reading someone's knowledge, you can't read their tractatii - because you already have their knowledge. Someone teaching you allows you to gain your own understanding of the subject from the teacher, so tractatii are still useful. This supernatural ability is faster, but that teacher can't give you anything new unless they increase their ability first.

Hi OneShot, I agree that hermetic magic can't do this (right now). But there are lots of supernatural virtues that do things that Hermetic Magic can't do or can't do easily. If this virtue is too overpowered, we can make it a major virtue. I agree that characters just generated can have high scores in an ability, and that is something to carefully consider. I suppose I wasn't too clear about the origin of this virtue. I see it as something that a character is initiated into. The society/cult/fraternity would have resources to teach people high levels of LM in order to make it easier to spread knowledge. Maybe we can add a flaw or two in the initiation script so that it makes it harder for them to learn the traditional way. They become used to being spoon-fed knowledge and lose part of the ability to learn.

As LM is actually knowledge of a language, I would argue that puissant can't apply to it. You do not know those symbols so you can't use them. Affinity with... is OK because you learn it faster.

If you make sure that your society/cult/fraternity is the only source of this Supernatural Ability, and no player character starts already as an initiate of it, you can use its initiation script(s) to balance the Supernatural Ability's overall impact in your saga. Otherwise you can't: the Flaws from the Ordeals just get baked into the character at creation time.

Puissant not applying to languages would be your house rule. If the Language of Memory shall be a real language, you might look up Ancient Magic (AM) p.10ff The Language of Adam, especailly p.28f Conclusions and p.29 box Magic in Other Languages, for some likely consequences of introducing a superior supernatural language into Mythic Europe.

The relevant bit of the Calc run description is fairly short, so I'll just quote the full thing here:

"This rune can be used in a rune script to create a repository of knowledge on a subject governed by the target rune. Thus Calc (Raido) could act as a repository on Area Lore, and Calc (Perth) could store Magic Lore or even Rune Magic. Once the repository has been created it must be filled by the caster; this process is identical to a character writing a summa or tractatus (ArM5, page 165), including the time taken. A tractatus written in this manner counts toward the maximum number of tractatus that an author can write. The completed runescript looks no different to any other at first glance, but if studied
reveals the presence of hidden depths, unfolding in the mind as if staring into murky water. Anyone in possession of the repository can study the knowledge contained therein, as if it were a text of the appropriate type, as long as they are capable of reading runes. Calc rune scripts can also be used to preserve smaller amounts of knowledge, such as a hero’s story imparted to anyone touching his grave marker."

And the relevant guidelines are:

General: The target receives a bonus or penalty equal to the magnitude of this effect when attempting to remember something, or when trying to recover information that has been lost
Level 5: Create a trivial memory in the mind of the target, equivalent to a few minutes of speech.
Level 10: Create a minor memory in the mind of the target, equivalent to a few hours of speech.
Level 20: Create a significant repository of lore filled and studied like a tractatus.
Level 30: Create a major repository of lore filled and studied like a summa

I realised myself by the end of my last post that what I was suggesting didn't really fit the description in your OP about what the ability was meant to do, but I'd already written it so figured might as well post it.

If you really want to keep the "writing down someone's complete knowledge about something" aspect, maybe the balance should be based around the mundane equivalent, that is teaching, in terms of xp gain and time. A really simple way to balance that would be that when you create a Language of Memory inscription (or whatever you want to call it) it has to be "taught" what the author knows, but using LM in place of teaching as an ability. Then it can be studied by anyone with the same or higher LM ability as if they were being taught by the author at the time of writing.

To use the example of Quintus again, with LM 9 he would have to write the book by "teaching" it until it reaches his teaching gain limit. He has Ignem 24 and his teaching (using LM instead of the teaching ability) grants -1+9+3 = 11xp per season so it would take him something like 28 seasons to fill the book up to level 24 Ignem.

Later another magus with LM 9 comes into possession of Quintus' book. He can study it each season and gains 11xp, exactly as if Quintus had been teaching him, and up to the 24 that would have been Quintus' teaching gain limit at the time he started writing the book. His apprentice only has LM 3 and can't read the book, however.

You could also allow a LM expert to intentionally write to be comprehensible to lower levels of LM. Quintus could have chosen to write the book at LM 3, but then the xp gained each season from the book would only be 5xp, not 11xp.

Or alternatively you could allow anyone to read any LM text (so long as they have at least 1 in LM) but they only get the benefit of their own LM ability on the teaching xp gain. So that apprentice of 3 LM could read the same book as his master, but only gains 5xp per season as he can't fully grasp the intricacies of the text.

That would basically make LM a laborious alternative to teaching, but with the benefit of producing a book which can be copied and potentially outlive the author.

1 Like

Hi OneShot,

If the Language of Memory shall be a real language

(I'm still learning how to use the forum tools)

No, the Language of Memory won't be a real language. It can't be spoken and it is hard to add to the language (all the symbols must be known by both author and reader. It is essentially a "dead" language because all of the symbols were created by the inventor).

I did have some ideas about it today, but it is a bit of a different direction than I was thinking before.

Today I thought that LM could be used with the Art of Memory from TMRE. Instead of a magus writing down their thoughts using LM directly, they meditate on their knowledge for a period of time to create a locus in their memory palace. At the end of their meditation the locus contains one or more texts written in LM about the specific memories. The magus can "copy" this into the real world using the rules for copying texts.

The meditation would have a series of ease factors based on the memory being translated.
I was thinking of using
Int + LM + stress die

Botch - each botch introduces inaccuracies in the memories, but the memory seems real.

fail by 10 or more meditate for the normal time but don't create a locus.
fail by 7-9 the meditation takes four times as long
fail by 4-6 the meditation takes three times as long
Fail by 1-3 the meditation takes twice as long
pass - the meditation takes the normal time
exceed by 3-5 the meditation takes 4/5th of the time.
exceed by 6-8 the meditation takes 3/5th of the time
exceed by 9-11 the meditation takes 2/5th of the time
exceed by 12 or more it takes 1/5th of the time

The meditation time is important because the magus can't eat or drink during meditation. As they are in a state of meditation they need to make deprivation checks half as often. e.g. water every 2 days instead of every day.
Meditating for 7 days would need 3 water deprivation checks and 2 food deprivation checks.

EF 3 momentary memory (1 page, takes 2 minutes)
EF 6 diameter memory (6 pages, takes 1 hour)
EF 9 hour long memory (30 pages, takes 10 hours)
EF 12 day long memory (50 pages, takes 2 days)
EF 15 text on an ability level (1 tractatus, takes 7 days)

The result of the meditation will occupy a locus in the creator's Memory Palace.
The number of pages or texts created will depend on what is being transcribed - each "text" will correspond to 1 level of an ability. Authors must always write the texts to their full ability (ignoring puissant). The texts are incomplete (useless) until the final volume is created.

This means that a person with ability level 6 will need to meditate 6 times with an EF of 15 to generate 6 texts. They cannot decide to only generate 3 or 4 texts.

Once these texts have been created in a memory locus they can be transcribed onto parchment using the standard rules. Each text takes a season to copy. A scribe wishing to copy these texts needs LM 1 to make sure that there are no errors.

Anyone can use the Art of Memory or Hermetic magic to copy these texts into a locus in their memory (1 text per locus), however:
One, they must have all of the texts in a series before they start to comprehend them.
Two, they must have at least LM 1 to understand the texts. They then enter into meditation to unlock and comprehend the knowledge in the texts. This is a seasonal activity which doesn't need a lab (but they need to be free from interruptions). Each day they enter into meditation to comprehend the texts in order. As each text is comprehended the pages fall become unbound and fall onto the floor of the locus. When the last text has been comprehended the pages scatter into dust and spread through the mind of the magus. The magus now has this knowledge and can use it freely.
If a reader already has a score in the ability being learned, the higher of the two is now used.

Notes: If the texts are read out of order the magus becomes confused, all read pages turn to dust and the magus has to try again another season.
An Int + LM + stress die roll can be made against twice the number of texts to check that the order written on the cover matches the symbols in the books.
Botch - for each botch the magus is convinced that 2 books are in the wrong position
fail - The magus doesn't know
pass - the magus can tell which order the books belong in.

After a week of texts not being comprehended the pages on the floor of the locus will turn to dust. The season is lost and all previous comprehension is lost. The magus will need to create new memories of the lost texts and try again.

For lower levels of abilites it is simpler for a magus to read a good summae. Even for mid-levels, assuming that a character already has a score in the ability it would be more useful to read tractati.
Most magi cannot spend 2 or 3 years without any interruptions.

Hi Argentius,
Many thanks for your ideas. The idea of using the memories as an alternative to teaching is an interesting one, but it seems like a lot of effort to use a minor virtue. Although being able to write a text to twice the normal limit could be considered worth a virtue.