Mnemonics based on sentence structure.

General Memory Chat
21

Hi Mnemoriam,

Like you, I too been tied up with commitments. Although most of mine have been of the antisocial and unprofessional variety. :slight_smile:

Actually, I was off work when I initiated this discussion, but have returned to the office, to family and to the usual domestic and other responsibilities, so time to focus is a a bit scarcer.iu

That said, I am still following this thread with great interest. I have been going through the links you most recently provided, reviewing the tags, etc.

As a result of exploring the various tags, I have been munching thoughtfully on the notion of clauses in general and subordinate clauses in particular. Right now I am seeing subordinate clauses and subordinating conjunctions as possibly pivotal structural elements to focus my attention on. When I am reading any material, I have started trying to train my mind to quickly spot the clauses (main clause, subordinate clauses, etc) and to evaluate them for main-eventness, or center-of-gravitiness, which ever term you prefer.

It has been awhile since I formally studied grammar, and at the time I had no particular purpose in mind, so I am having to reeducate myself and test out whether my intuitions make sense in practice.

One thing I have noticed is that, even though subordinate clauses are supposed to represent less important info that the main clause, or the clause to which it is subordinated, that is definitely not always the case, as our discussion around play sentence 1 showed. Importance depends on one’s objectives in the moment–for our purposes, anyway.

At this moment, I am supposed to be getting washed up for work. If I don’t hurry, I will have to make do with a standard of hygiene reduced by about half the recommended standard, so for the sake of my workmates, I had better get going.

Please do keep commenting. I am looking forward to hearing your further thoughts, even if I can’t quickly respond.

Regards,

Darn

22

Hi Mnemoriam,

Just a note that I am still playing with ideas from this discussion. I have also been thinking about another thread in which you discussed the problem of words that take many forms. (https://artofmemory.com/forums/a-simple-question-for-a-probably-complex-answer)

In comment #7 of the linked post, you used an example involving the word hammer, hammering, hammered, but in a variety of forms, each with different semantic implications. Related to all this, I am toying with two concepts: sentences and stories.

What follows is merely a rambling record of some of my recent thoughts:

I am taking it as axiomatic that sentences are stories. At least that is how I want to think of them for mnemonic purposes. One of the most basic sentence structures is the subject-verb-object (SVO) sentence structure, at least for English speakers. SVO sentences also provide one of the simplest story structures. E.g., “Ann bit Charles”. It is also a very basic mnemonic structure.

As discussed previously, in the real world we tend to encounter sentences that are far more complicated than the simple SVO sentences that I am referring to above. They tend to accumulate additional parts, like adjectives, adverbs, phrases and clauses that modify the story in a variety of ways. Here are a few of the main ways:

  • Adjusting specificity
  • Clarifying history: prequel, context and sequel
  • Clarifying identity of actors, objects and other elements<\li>
  • Rationalizing: Clarifying cause or purpose<\li>
  • Clarifying locality: place and time<\li>
  • Elaborating<\li>
  • Describing<\li>
  • Clarifying directionality<\li>
  • Providing narration: E.g., "Sally murmurred, ashen faced"<\li>

All of the above sentence functions can also be considered story functions, and consequently mnemonic functions.

Getting back to your example with the word hammer and its manifold variations of form and meaning, I am thinking that each variation would prove to be a special case of a story. If we could look closely at the history that begat each new form, we could use it to tell a different story. I am way too lazy to embark on such an etymological safari, so I will be smugly satisfied to procede on the assumption that my conjecture is accurate.

I was thinking that it is possible to step through your example story of the two boys and the swordsmith finding ways to reference each variation of the word hammer in a way that is useful mnemonically, but preserves some semblance of the original semantic sense of each variation.

To do this, though, I found it necessary to insert storytelling devices like flashbacks and scenes in which some characters freeze others in time while they tell a story within a story.

The trouble with this approach is it is easier to do than to explain. Also, this process was made easier because the example story you gave was already fairly easy to visualize. If we were dealing with an abstract passage from some abstract philosophical or scientific writing, it might be a lot tougher.

Anyway, these are all ramblings. Not sure if there is anything you would think worth gleaning from this, except that I am still futzing around with the same general topic.

If I had to summarize, I might say my objective is to take what I read, eliminate the unnecessary, then convert the rest into the simplest possible sentence, using SVO as my archetype, and thereby construct simple memorable stories that parallel the structure of the information I want to store and recall. The various functions, as listed above, that complicate the sentence can be treated as part of the story. When I truly need to distinguish between variations in a word’s forms, I can use story-telling devices like freezing, flashbacks, or even flash forwards.

Best regards,

Darn

23

Hey tarnation!

Thanks for the message.

This is the briefest possible response, because I’ve just burned the few spare minutes my wife so generously allowed me before removing me from my beloved anti-social behaviour.

This is a perfect summary of my own intentions too. In fact, I wrote a huge post intended for this thread trying to put all I’ve been thinking about in some kind of structured form. However, although I began the post excusing myself for a “probably nonsensical post”, it ended up not passing my personal quality-control. I promise I’ll revise it, so that I too can contribute to this really interesting thread (not that anyone but you agree with me on the “interesting” portion of what I just said).

I totally agree with your focus on story-telling and, in fact, I have been thinking a lot about this. This thread, which I recently found, makes and interesting parallel between different approaches to linking and their relation to story-telling. I was thinking of coalescing the entire thread into a kind of taxonomical classification of the various approaches and related nomenclature. I think it would be an useful exercise. No time yet, though.

But I must confess I didn’t understand what you meant about using story-telling to disambiguate between word variations. I will reread your post as many times as necessary and I’ll also reread my own post about the “hammers”, then I’ll comment on your idea.

Well, as always, the “briefest possible response” got much bigger than I anticipated. Time to go smile for people I don’t know and won’t probably like. Thanks God they created the “single-malts”…

Best,
M.

P.S.: Have you ever heard of Ogden? He created Basic English and if your follow his website, there is a lot of info on how to write simple. In fact, there is a “Simple Wikipedia”, do you know that? I think there is a lot in there for our purposes. At least, a few insights… Check out, for instances, the “classes” of sentences they consider simple enough:

Subject-Verb-DirectObject.
Subject-Verb-IndirectObject.
Subject-Verb-DirectObject-IndirectObject.
Subject-Verb-DirectObject-SubordinateClause.
Subject-Verb-DirectObject-IndirectObject-SubordinateClause.

Just fuel for thought… Best!

24

I tried the ‘quality control’ thingy awhile back, but abandoned it, along with this thing people call ‘standards’. Doing this has greatly increased my ability to produce content. The only exception is that next week I plan to have my conscience surgically removed, which I am doing to improve the quality of my sleep.

On the question of disambiguating word variations, I didn’t actually explain, just talked around it. I think it would take some time to explain in writing, so I will have to come back with better explanations a bit later.

Indeed, I do know of Ogden and Basic English. I love simplified languages. (Also artificial languages.) aon the topic of simple English, my hero was Rudolph Flesch, who wrote books like the Art of Plain Talk. He would show how to simplify all kinds of high-fallootin-gobbledegook type sentences. He convinced me long ago that what passes for highly academic English in many circles nothing more than mush-brained bullsh*t obfuscation. If you use Word as a document editor, there is a function that ‘grades’ sentences for simplicity or complexity that is largely based on a calculation he designed.

As for your little list of simple sentence classes–yes, exactly the kind of structures that interest me. I boiled it down to SVO for simplicity of expectorating my sentential verluptuousity, but in the back of my mind, I was thinking that reality would require several simple variations of the kind you list.

Regards,

Darn

25

An additional point regarding objectives…

There is a concept that I have in my head that I might call “minimal encoding” or sometimes “half-encoding” and it relates to my goals.

Typically, when we encode information mnemonically, we end up with images that are very unrelated to the information they are helping to store. For example, we often encode an item based on a pun, or word-play of some kind, not on anything semantically related to the main idea we are trying to capture. This generally works well, but there is a cost in the time and effort to encode and the time and effort to decode. There is also, for me at least, often a sense of disconnect, or incoherency, between the stored images and the desired knowledge.

While experience helps me deal with my discomfort at this sense of disconnect, I would really like to have as much harmony and coherency between the stored images and the desired end-knowledge. To me, this means that the stored mnemonics should be minimally encoded–that is, they should be as close semantically to the end-product as possible.

I doubt it would be practical to achieve minimal encoding all of the time, but I suspect it might be possible for a body of info to reach half-encoding on average. In other words, we might often be able to minimally encode images, but at other times we would need to heavily encode our mnemonic images. (Image isn’t the precise word I want, since not all encoding is in images, but it’ll have to do for now.) The point is to come as close to the goal as is reasonably possible.

This effort to structured approach is, for me, all in the service of trying to work toward minimal encoding.

26

Hi tarnation,

I hope you know that I totally agree with that and it has been one of my main objectives to find ways to do that. I was rereading that “hammers” post you mentioned and I believe I have been very clear about that in post #3. So, we are pretty much on the same page here. However, practice has shown me that sometimes we can’t avoid using other methods beyond “minimal encoding” (I like to use simply “semantic encoding”) and I believe you will agree with that too.

I’ll use this post to clearly distinguish between the three main methods I believe we use for mnemonic encoding, namely, (i) sound resemblance, (ii) semantic (minimal) relatedness and (iii) free association. This will also greatly facilitate your understanding of my next post and probably foster more discussions.

Let me use a previous example of this very thread and your own encoding first:

TEXT: For who would bear the whips and scorns of time,
MNEMONIC: Four owls (hoo - who) giving birth to (bear-ing) whipped cream (whips) and scones (scorns) made of dimes (time)

For → Four: # sound resemblance
who → owls # free association
bear → giving birth (bear-ing) # sound resemblance
whips → whipped cream # sound resemblance
scorns → scones # sound resemblance
time → dimes # sound resemblance

It seems that, although you like the idea of using “minimal encoding”, you are quite biased towards sound resemblance, no?

Now, another example from Josh Cohen, now with his encoding:

TEXT: The pangs of dispriz’d love, the law‘s delay,
MNEMONIC: A heart (love) with sharp fangs (pangs) is holding a carnival prize (dispriz’d) and standing to the left of Judge Dredd (“I am the Law“) who is holding an hourglass (delay).

love → heart # semantic relatedness
pangs → fangs # sound resemblance
dispriz’d → carnival prize # sound resemblance
law → Judge Dredd # free association
delay → hourglass # semantic relatedness

I think this is all quite clear, I just wanted to set this straight upfront and see if you agree with the taxonomy and if you have any other method to add to the list.

Now, the main post…

Best,
M.

27

This simple summary of yours (which I agree with entirely) poses huge problems, which are the ones I am trying to tackle, namely:

  1. how to extract the gist of texts;
  2. how to simplify that gist;
  3. how to encode the resultant sentences;
  4. how to store them.

This can all be seen as a matter of simply reading well and applying usual mnemonic techniques. However, I’d like this process to be as well-defined as possible, so that it can be explained and reproduced with precision. For that, I believe that the best way to achieve such level of explanation, reproducibility and precision is to turn the entire process into a working computer program. I don’t mean necessarily implementing everything, but at least, to think of the problem as such.

This is my attempt to make clear my “mission statement” here. I believe that at least portions of it resonate with your own interests, thus this ongoing thread.

That said, for the sake of sanity, I’ll focus right now “only" on number 3 above.

Let’s simplify the problem even further and assume we only have SVO sentences. I totally agree with your comment that sentences “tend to accumulate additional parts” and with all the “functions” you listed, but let’s start simple.

How to encode such sentences? This is the tentative pseudocode I’ve come up with:

pseudocode.png
pseudocode.png263×600 77.2 KB

Trying to explain this pseudocode as simply as possible:

The main function is encode and basically what it does is to first check if the word is concrete. If so, it checks if it is a general word (probably this only applies to nouns, but I am not so sure): for instance, if the word is “car”, try to specify it to a Ferrari 488 GTB or whatever. Then, directly encode it, i.e., use the specific word as an image. If the word is not concrete (democracy, perceive, etc.), find an appropriate substitute word. If it is not a noun, first turn it into one, i.e., nominalise it (e.g. “perceive” becomes “perception”).

To find substitute words, there are those three options I mentioned in the previous post:

1 - Sound resemblance;
2 - Semantic relatedness (minimal encoding);
3 - Free association.

Then, because it is a simple SVO sentence (our initial assumption), we can simply link the first image to the next and then to the next, providing the final encoded “story”.

If you notice, this is a compromise (mixed) approach with respect to my previous idea (stated on post #6 of this thread) that all words should be turned into nouns. Now, only if it is not concrete we’ll do it. For instance, if the verb is “to kick” then we simply imagine the action of kicking. There is no point in turning this into a pair of “soccer boots” or whatever. Adjectives, in turn, will always have to be converted to nouns. As an example, check the list below. I doubt there is another way but to turn them into nouns.
adjectives.png

Also, abstract verbs such as “like” or “perceive” will have to be converted to nouns. Then we simply link everything. I don’t want to delve into sentence structure right now, but the use of adjectives already hint into the problems we’ll have. If the “O” of the SVO contains an adjective or if the “S” is two people, for instance, we’ll have more than three images connected to each other and now we’ll have to enforce sentence structure somehow, lest we might lose important information from the original sentence.

Well, I’ll stop for now and work on a more complex case. Please, analyse this post with care and tell me if you think we could use this simple pseudocode as the basis for our discussions from now on. Of course, make all the comments you want and propose changes as you wish or even deny it completely. My point here is to try to find some common ground, as objective as possible, so that we can try to actually generate palpable results from this great thread.

Thanks for keeping the discussion alive.

All my best,
M.

P.S.: Both semantic relatedness and concreteness index can be calculated using WordNet. A very interesting paper on simulating human concreteness ratings based on WordNet is found here.

28

Love it! I will try doing some step-thru’s with the pseudocode to see how it works for me.

[ASIDE: Going back to the whips and scorns of time example (#26 above), it is true that I am biased towards sound resemblance and wordplay in general, but the intent was to provide an example of keeping the structures more or less parallel between the source and a (heavily) encoded material. This was in keeping with my original intent when I started this post. I only say that as an opportunity to call out (or confess) that my objectives have changed to a fair degree thanks to our discussion. I still think my original idea is good: maintaining a parallel structure is an aspect of minimal encoding, but it’s definitely not the only thing. Although I haven’t abandoned the parallel structure objective, I have decided that the more important goal is to maximize minimal encoding however it can be practically achieved. Your notion of semantic encoding seems a fair bit more evolved than my structure idea, and has lead me to want to explore much broader goals than I originally envisioned. NOTE TO FRIENDS AND STRANGERS: if you are reading this at some time in the future because I have done something heretical or crazy, you can probably point to this discussion as the ground zero of my radicalization. :D]

29

Hi,

I found an interesting book reference that might formalise a possible path for our objectives. It is called “Rhetorical grammar: Grammatical choices, rhetorical effects”, by Martha Kolln. I haven’t bought it yet, because I am not in the mood to spend 70+ dollars in a 300- page book right now, but I might in the future if I don’t find a better/cheaper reference.

For now, this link provides what matters most for us, which are seven basic sentence structure patterns that, according to this other link, covers 95% of all English sentences. I am sure we can convert pretty much all sentences to fit these patterns with minimal loss, assuming we do some breaking-up into main clauses first.

If we could create seven very distinct visual patterns for mental images, then we could retain original sentence structure just by the visual composition of our images. This is the idea I already had, but now we at least have a formal definition of main patterns to use. I still don’t know how to do it (and even if it would ever work), but it is a start.

The article also provides the main “glossary” we need to think of, such as “linking verbs” (a term I did not know of) and subject and object complements (terms I know of, but am not comfortable with). The problem, as before, is that complements can be entire phrases with multiple words, so, in theory, there would have to be “nesting” of images in some of those seven patterns, complicating everything.

Well, no one said it would be easy…

Best,
M.

30

Hi Mnemoriam,

We might imagine a few imaginary building templates with 2 to 4 basic rooms:

Building template #1 (Be pattern 1):
Room 1. Subject
Room 2. Be Verb
Room 3. Adverbial

Building template #2 (Be pattern 2):
Room 1. Subject
Room 2. Be Verb
Room 3. Subject Complement

Building template #3 (Linking verb pattern):
Room 1. Subject
Room 2. Linking Verb
Room 3. Subject Complement

Building template #4 (Intransitive verb pattern):
Room 1. Subject
Room 2. Intransitive Verb

Building type 5 (Basic transitive verb pattern):
Room 1. Subject
Room 2. Transitive Verb
Room 3. Direct Object

Building template #6 (transitive verb + two complements - type 1):
Room 1. Subject
Room 2. Transitive Verb
Room 3. Direct Object
Room 4. Indirect Object
[Note: Rooms 3 and 4 could be reversed, if you prefer. I prefer this order for consistency with type 7.]

Building template #7 (transitive verb + two complements - type 1):
Room 1. Subject
Room 2. Transitive Verb
Room 3. Direct Object
Room 4. Object Complement

These building types could be compressed further by combining sentence types 1, 2, 3 and 5 into one building type and 6 and 7 into another. This would require only 3 basic building templates:

Sentence type 4 would use a 2-room base template.
Sentence types 1, 2, 3 and 5 would used the same 3-room base template.
Sentence type 6 and 7 would use a 4 room base template.

Personally, I think I would stick with 7 separate templates for now and decide on further compression later.

While these 7 correspond to the basic patterns, there is still a need to account for complexity caused by additional clauses, I think I would start with the basic rooms from the template, hopefully having very distinct qualities, then add on rooms wherever the non-basic clauses are needed. In some cases, if these additional clauses didn’t seem important for my purposes, I would just leave them out.

PLAY SENTENCE #2

“Scientists from the Icahn School of Medicine at Mount Sinai, the University of California, San Francisco, and Harvard Medical School used a rigorous study design to assess the biological impact of meditation compared to vacation.”

Excerpted from: Systems biology research study reveals benefits of vacation, meditation | ScienceDaily

Comments:
This looks to me like sentence pattern 5–the basic transitive verb pattern, but with an additional “to” clause that indicates the purpose of their actions. So, it would start with building template #5:

Building template #5 (Basic transitive verb pattern):
Room 1. Subject - Scientists (from the …)
Room 2. Transitive Verb - Used
Room 3. Direct Object - Design (a rigorous study,)

I think I would create each template so it has a distinct feel to it. Perhaps each having unique styles of furnishings and subject rooms would be very different from verb rooms, which would be distinct in some way from direct object rooms, and so on.

I would start by isolating a central word image for the subject phrase, a central word image for the verb phrase and a central word image for the direct object, or other 3rd and 4th elements. Then I would start adding the remaining details in each room. So, in my subject room, I might start by picturing Bill Nye and Neil DeGrasse Tyson working around a main table in a central position in the subject room, dressed in white smocks and looking through microscopes, and generally doing something science-y. Then around the room, I would add images to capture the remaining details of the subject–for example, the room might have coloured glass windows like you see at a church to remind me of icons (Icahn).

In the next room–the activity room, in my mind–I might place a man taking drugs through a bong to represent the verb “to use”. Then in the direct object room, the central image might focus on some artistic designs and some artist doing a study of a body in rigour mortis (rigorous study).

After filling the base rooms with images, the next step would be to construct an additional room on the fly to handle the clause that expresses purpose: To (assess the biological…). This room might have the feel of an extension to the main building–something to set it apart from the 3 base rooms, like a basement or a garage.

Take care,

Darn

31

Continuing with the above example, I might handle the additional clause–the “(in order) to” clause–by imagining a corridor leading from the 3 main rooms to a set of stairs leading up to a second floor where there would be a main room and a hidden room. The main activity room upstairs would have a glass mirror, behind which I know the scientists from the subject room are hiding as they carefully observe events in the main room. In the main room, there might be a crocodile playing soccer with Spock (bayou-logical) on one dish of a large balance-scale and a meditating yogi on the other dish. To me, the scale represents assaying, which I connect to the verb assess.

32

From all of this, I should be able to reconstruct the idea of the sentence as follows:
“Scientists” + specifying details + “use” + (rigorous study) design + in order to + assess + biological + impact + [compared to] + meditation.

(I forgot to mention in the previous comment above that Impact is the name of a soccer (football) team in Montreal.)

There was some non-minimal encoding in all this, but not too much, I think. Especially considering that and encodings like crocodile & Spock, automatically evokes bayou-logical, thence biological. This is a encoding I would use very consistently, so it is like the word itself in my mind. Thus, it does not feel heavily encoded to me, even if it might to an outside observer.

33

As I think about the structure of the building, perhaps I would change something. Not sure if my reasons will shine through. Perhaps it will just seem confused, but rather than have a corridor that leads from one room to another, I might imagine that there are doors leading directly from room 1 to room 2 to room 3–that is from the subject room, to the verb room to the direct object room. Then, when I need to add an additional room for the “(in order) to” clause, it also needs to emanate directly from the verb room, but should also have a connection to the direct object room. I might make it distinct by creating a passage–as opposed to an ordinary doorway–that is leads out of the verb room into the area that I mentioned earlier with the crocodile, etc. The passage might still involve a small stairway. The observation room then might have a passageway in from the verb room and another passage out to the direct object room, to which it also should be connected. Not sure if I would still want the hidden observation area.

Since the scientists from the subject room are responsible for all the action in this sentence, I would imagine them walking from their subject room, through the verb room, where they would observe the drug user. From there, they have a choice: they could either go through the door into the direct object room, where they would see the person with the designs doing the artistic study of a body in rigour mortis; or, they could go into/up the passageway/stairs to see the “(in order) to” room where they see crocodile, etc. Whichever choice they make, they can pass now into the remaining room to get the complete meaning of the sentence.

It sounds complicated, but it was actually fairly easy to construct. Not sure how it would work if I were doing 20 such sentences in a row, however.

34

Just a quick update.

Using these sentence archetypes seems to be working very well for me. They also mesh with another project that I have been working on, which is the creation of loci-on-the-fly–that is, virtual loci that are created based on the information itself.

I have already been doing something similar based on letters in a word and I find they seem to be very good for retention. I suspect that is because of the extra mental effort they require and because the loci relate directly to the information. Now, I have at least two methods, one based on letters and another based on sentence structure. I suspect merging the two techniques will prove very useful, but I need to play with this more.

On a separate note, I have also done some step-throughs of your (Memoriam’s) pseudo-code and it seems very natural.

35

Hey tarnation,

Sorry for the absence. And I am glad to see you practicing with these new ideas.

Your last sentence perfectly summarises my main concern with your approach. It does a great job for visualising the given sentence: it kept really well sentence structure and meaning. However, if I had to give you a confidence interval of the number of one-sentence-notes I produce out of a decent book, I’d say with 95% certainty that it has an average of 30 sentences plus or minus 20. How could your system scale to encompass 50, 30 or even 10 sentences? What to say of hundreds of books?

Your imagination is no doubt much greater than mine, but even so, I think that changing the “theme” or shape of objects or color or whatever of a fixed structure of rooms will not suffice. What are your thoughts on coping with this?

Your idea resembles, in a way, a post from this forum by r30, which I have always found really interesting, but utterly impossible to fully understand. He uses his system for “memorising things on the spot”, so I don’t believe he meant to use it for permanent knowledge, but more for daily stuff such as appointments, grocery lists, etc. This is my interpretation only, simply because I think his system has the same scalability problems as yours. It would be interesting if he could provide new insights about the usefulness of his system; and maybe you could take a look at his post to see if something resonates with yours.

I had written the above portion of this post before your last message; So, I am glad to hear that. How exactly have you been doing to differentiate between archetypes of a single pattern? For instance, if you have three sentences of one single pattern, how do you differentiate the rooms?

This quote of yours seems to really resonate with r30’s post I mentioned. I think you should check it out. So, I believe you are making the images themselves “become” the loci or something like this, right? I’ve always wanted to do something similar, but I haven’t been capable so far. I’d be very interested to hear about your progress on this technique.

With respect to my progress on all this, I am actually quite excited right now. As Fijordan “accused” me, I am definitely “thinking much more than practicing”, but I am enjoying it a lot. For a long time now, I have had two books of John Sowa, a renowned computer scientist, in my wish list. He created “conceptual graphs”, a formal language for knowledge representation through graphs. I’ve always thought that could help me somehow, but only now I could visualise the potential with more clarity. As I read about all these subjects that I deem related with our memorisation purposes, I slowly understand better my own idiosyncrasies: Why am I so crazy about this “systematisation” of mnemonics? I am beginning to understand (or to accept) that I always crave for finding logical ways of doing things – I need pattern. And what these guys do in the realm of artificial intelligence/knowledge representation (the way I see it) is to reduce the workings of the world into their constituent parts, finding the patterns that arise from them and formalising them into a precise language. In doing so, they allow further reuse of this “processed information” in countless ways. Their main aim is for computers to use this information, but I argue that there are many uses for humans too.

If you have some spare time, check this paper out. Please, don’t read it fully; but I think you’ll better understand where I am getting at. By using a formal language to process an “informal” one (such as natural language), we reduce it to its forming patterns – we have to, otherwise, there will be too many gaps in the end-result. That’s exactly what we are trying to do here. I think it would be a huge help to have a program assisting us into coding (like you, I like to coin new terms and I usually refer to that as “mnemocoding”) original sentences into a complex image, or a sequence of images such as using your system of adjoining rooms. I believe conceptual graphs may show the way, so I finally bought the two books I mentioned.

What I am currently doing before the books arrive (it will take some time), is trying to understand conceptual graphs better through this online course and trying to gauge what the state-of-the-art is. It seems that conceptual graphs are not being used much nowadays and “Formal Concept Analysis” have largely replaced it, and I want to know why.

Well, I guess we are both slowly getting somewhere – you, in practice, I, in theory – hopefully, we will eventually meet at some point.

Best,
M.

36

Hi Mnemoriam,

I checked out r30’s post. It definitely interests me, but I must admit, I was having a hard time grasping his meanings. There were a few things that I walked away with, but wasn’t sure if I was understanding his intention, or just filling in notions rorschach-like. I am intrigued by his plan, but will have to revisit on a vitamin-intensive day when I am ready to engage my brain more energetically.

My plan was not that the image become the loci, although I do resort to similar techniques, if I understood correctly. I would prefer to say the objective is to have the “information become the loci”. Hopefully, the distinction will become clear.

My letter-based system depends on some alphabetized lists I have been building of all kinds of things that could be used to construct various loci-on-the-fly: types of buildings, types of outdoor spaces, types of rooms, types of passages (e.g.: stairways, corridors, alleyways, walkways, canals, bridges), types of furniture, types of fixtures, elements of landscape, objects of art, elements of interior design, vehicles, animals that might seem appropriate as background images, and so on.

Then I look for a small number of key words that capture the essence of the thing I am trying to memorize. Typically, this involves a title or index word, which is a word that has high value as a starting point for recalling the rest of the info.

So, when I wanted to memorize the cranial nerves, I chose the words “CRANIAL NERVES” as my starting point.

The next steps don’t follow any fixed rules, other than to construct a loci from the letters of the word, drawing on the aforementioned alphabetized list for inspiration.

For my cranial nerves example, I created an overall place using the imaginary image of a gigantic (think of a 3 story house) skull with a very tall crane (long-legged bird) standing by it, figuring the redundancy would help reinforce the idea of a cranium or the word cranial.

The skull, is actually a location, with the orifices being rooms. imagining it to be the size of a small apartment building, I focused on the room that made up the hollow of the right eye. There I placed a thing from my list that started with C, namely a CAT. I might easily have placed other C objects, for redundancy, such as a CAT on COUCH (we also used to call a sofa a Chesterfield, when I was young), or some such thing. My rules are pretty loose, and I am constantly changing them. In the left eye, I coukd make out a ROCKING CHAIR, representing the letter R of the word ‘cranium’. Next, I had an ASP passing through the skull from the position around where the right ears would be, through to the left ear. The skull didn’t seem to have enough rooms, for my needs, but no matter. i added a NEST at the foot of the crane and an IRON GATE in front of the skull-shaped apartment building, then an ALLEYWAY to the left side of the skull-structure. Further to the left was a LIBRARY building, then a large NAVAL SHIP with an ESCALATOR and a separate RAMP, then finally a person on the dock driving a vehicle with a large VACUUM.

Thus, In addition to the crane and the skull itself, I had a palace that included the following loci:
Cat
Rocker
Asp
Nest
Iron gate
Alleyway
Library

Naval ship
Escalator
Ramp
Vacuum

I am sure I don’t need to repeat that the initial letters spell CRANIAL NERV

For me, this virtual palace is easy to visualize and extremely unique. With the palace constructed, it was just a matter of using these loci to place the names of the nerves:

Crane - Olfactory

Cat - Optic
Rocker - Oculomotor
Asp - Trochlear
Nest - Trigeminal
Iron gate - Abducens
Alleyway - Facial
Library - Vestibulocochlear

Naval ship - Glossopharyngeal
Escalator - Vagus
Ramp - Spinal Accessory
Vacuum - Hypoglossal

[Forgive me if I don’t give the specific images I use to link the nerve names to the loci. I figure anybody reading this far would not benefit from those details.]

In my experience, memories stored using this method have a fairly high persistence rate, requiring less repetition practice. I think that is because it involves in a dedicated memory palace and the requires a fair amount of creative energy.

Currently, I find the method a bit slow, but I think I can improve my speed with practice.

What this description might not illustrate as emphatically and effectively as I would like, is the idea that it is possible to build palaces, rooms and spaces in a way that is both creative and aesthetically pleasing, and the very act of trying to recall the information reminds you of the palace, which in turn helps you find the details.

37

The above method may be far from the original topic, but it does seem to result in highly unique, highly memorable artificial loci, so should be fairly scalable, as long as I do not need to reuse the index words (cranial nerve, in this example) over and over.

As far as how all this relates to the 7 sentence archetypes and the scalability challenges you raised, I am still ponder-if-icating how I might best go about addressing that. At the very least, I believe it is possible to manage some reuse of the archetypes. I think using the (currently alphabetical) list of loci ‘parts’ for inspiration would make it not-too to improvise a number of unique palaces, but some reengineering might be needed.

38

We have this in common. I am glad you are bent this way because I rarely get the opportunity to talk to other people who are interested in such things. Perhaps we could get nearby cells when they come to put us in the rubber rooms.

Actually, if you have any interest in this sort of thing, you might want to look into your Myers Briggs type. I was long ago told by a professional psychometrician that I am an INTJ, which is slightly rare. I didn’t make anything of it at first, but I later realized it was a fairly good description of my nature. Among other things, INTJs are supposed to be a bit obsessed with systems. Not sure about the scientific underpinnings of the Myers Briggs method of analysis, but I would say it is better than horoscopes, and seems to be very good at describing me.

39

Hi tarnation,

I think your letter-based method really interesting and intriguing. That’s the kind of thing I aways think about, but never commit to for the following reason: it does not make use of the spatial awareness so characteristic of the method of loci. At least, not nearly as much as using a “proper” memory palace, a real place you are acquainted with.

We are digressing a bit from the previous topic, but this is a discussion central to my objectives as well. This kind of system has great scalability potential, as you mentioned, but so does the linking method or even the peg method. What I want is a loci-based scalable method, which yours sort of is, but I wonder how effective it is compared to a “real” loci-based method. I hope you get what I am saying, I am in a kind of a hurry here, so I don’t know if I am expressing me really well.

I am very interested in understanding your experiences with respect to this letter-based method. I don’t know if you remember somewhere in our thread when I mentioned this crazy idea of using the keyword of the main event as the memory palace per se, and for that I would visualize the object in three dimensions as a topographic image. The whole reason for the idea is to (i) provide a scalable method and (ii) try to generate as much spatial awareness as possible. If I can visualise the object as a real 3D terrain, I might eventually have the feeling of having been there, just like playing using a video-game-based memory palace. However, I also don’t think video-games are nearly as potent as real memory palaces, but they are better than simply imagining a place based on objects.

So, I ask: how well can you actually feel inside your letter-based memory palace? Are you really making use of our brain’s capacity for spatial awareness and orientation, or does your method work more like a peg method, where the linking occurs as if floating in front of our mind’s eye, not actually placed in a spatial location/loci?

Sorry, this time I’ll have to forget about my “quality control”, so I really hope I making any sense.

Thanks a lot for any further clarifications.

Best,
M.

40

I have never been really interested in this kind of thing, but I googled it after you mentioned and ended up doing two free online similar tests. I must say I am impressed.

The first one I did is very simple and it provided the following result:

30% INTJ
15% ISTP
14% ISTJ
12% INTP
11% INFJ

The second one was much more elaborate and the final result really impressed me. The explanations match so much of myself that I got almost irritated for being so normal as to be represented by a formula. Here is the result:

Your personality type: “The Logician” (INTP-T)
Strength of individual traits: Introverted: 59%, Intuitive: 75%, Thinking: 58%, Prospecting: 60%, Turbulent: 95%.
Role: Analyst
Strategy: Constant Improvement

I guess we do “have this is common”…

Best,
M.