I found it pretty curious that the voiced dental fricative was used to represent 1 but the unvoiced dental fricative was used to represent 8, seems like the only voiced/unvoiced pair to be split in this way.
In my idiolect, they both belong together with the alveolar plosive pair (/t/,/d/) in category 1. The place of articulation feels almost identical between my laminal plosives and dental fricatives.
Though I do agree with grouping the velar nasal along with the alveolar nasal rather than with the other velar plosives. The acoustic similarity between /n/ and /N/ feels a lot more salient than any similarity in velar articulation.
Overall, a really cool system! Now I'm curious about what kind of density you could reasonably encode if you remove the need to memorize and used it as a form of steganography.
You're right that the voiced and unvoiced pair ð and θ probably belong together, as is the case with other voiced and unvoiced pairs. This would also significantly reduce the disambiguities that need to be manually resolved in Step 3. I chose to put θ into category 8 though because intuitively it sounds more like f than t to me, possibly because I'm not a native speaker. I might change my mind about it if it turns out to be more practical to have them both in category 1.
I like your idea of using the major system as a form of steganography. It would require a deterministic encoding algorithm though. I might explore this idea in another post in the future.
I think these are cases where it's going to depend on what is most natural for each individual, especially when used for mnemonic purposes. Whether the acoustic or articulatory similarities feel more salient is going to depend entirely on your own mental model of the language. It's common to see the unvoiced dental fricative realized as /f/, even in native english dialects, so if that is more natural for you to remember, there's value in that.
It might or might not be easier for people to start by reading the Wikipedia page about this system that maps digits to sounds (which has a huge history and fanbase):
Slightly off topic and I get your point but I cracked up at myself when I read "maps digits to sounds" and my first thought was "oh like words like seven"
Mnemonics are vastly overrated and the tricks you see with people using them are mainly a result of constant practice which could be achieved by other means.
"Gruneberg (1998) argues that the keyword method, in general, is inferior to rote learning in the longer-term retention of vocabulary."
Interesting article! I agree that mnemonics don't help you much with long-term retention, that's why it makes sense to combine it with a spaced-repetition strategy such as rote learning. However, mnemonics help you translate information into a form that's easier to process for the brain. Even if you don't actively use mnemonic techniques when memorizing meaningless information, you will likely end up applying such techniques intuitively.
Interesting; I learned a slightly different set of associations with mnemonics for the sounds as well. 0, 2, and 3 were the same as this system:
0 = the word zero has an s/z sound
1 = a single line, like lower-case "L"
2 = two lines, like lower-case "N"
3 = three lines, like lower-case "M"
4 = the word "four" ends in an "R" sound
5 = the word "five" contains "F" and "V"
6 = the digit looks like a lower-case "B", and lower-case "D" is its mirror image
7 = the digit looks like a badly-written upper-case "T"
8 = the "gh" in the word is weird, and reminiscent of "CH"/"SH" (and "J" is similar)
9 = the digit looks like a lower-case "G", which is similar to lower-case "Q", and lower-case "P" is its mirror image
I had no idea that there was a standard, but I use it a lot to remember what page I've reached in a book (without a bookmark handy), or for short strings of numbers like IP addresses. I find it much easier to remember "Latin insults suns lamely" than "172.217.20.131", though I'd probably try and find a better mnemonic than "insults" for 217, as that technically maps to 20170...
It's kind of like the "correct horse battery staple" thing (https://xkcd.com/936/); words are easier to memorise, especially if you can create a narrative, however tiny.
This is great. I started hacking on something very similar (https://ezmonic.com) but you took it all the way with NLTK and finding noun phrases. Another great addition would be to find rhyming phrases if it's possible given a number. Maybe add some rules to the Mnemonic system so that certain parts of speech are ignored, so you can throw extra syllables in when needed by adding some adjectives or adverbs. I'm curious if the flexibility to create rhythmic rhymes would aid in memorization, or if the added phonemes would cancel out the benefit.
Great idea about finding or even generating rhyming phrases! If you clearly define the parts of speech used for encoding (probably nouns), I don't think it would cancel out the benefit.
Maybe it's just me, but reflecting on old memories has shown me how unreliable the human memory is, so nowadays I just try to rely as little as possible in my own memory.
I also find it takes a "computational" toll on my thinking ability. As far as I'm concerned, my mind pretty much has only volatile storage, and anything that should be kept for later needs to be written back to something more durable.
The post is not uninteresting but I doubt it has any real life applicability. Memorizing the number->consonants mapping (which is lossy) probably take already more time and efforts than memorizing the few raw digit strings that much be memorized in one’s life. It also assumes people would recall accurately a phrase or sentence.
My personal technic for memorizing numbers like pin or phone numbers is to memorize the general shape of the difference between each digit. It’s obvious to compute, use visual memory and doesn’t not requires an external system.
I use Major for dates, and I know the periodic table as a mapping of number <-> element <-> symbol. (And, by accident, I know my Oyster card's serial number.) It also is handy if you want to carry a number around briefly without writing it down - I've used it for carrying numbers from computer to computer at work. With lots more practice I could probably get near-real-time memorisation, and the possibilities would expand much more.
In my idiolect, they both belong together with the alveolar plosive pair (/t/,/d/) in category 1. The place of articulation feels almost identical between my laminal plosives and dental fricatives.
Though I do agree with grouping the velar nasal along with the alveolar nasal rather than with the other velar plosives. The acoustic similarity between /n/ and /N/ feels a lot more salient than any similarity in velar articulation.
Overall, a really cool system! Now I'm curious about what kind of density you could reasonably encode if you remove the need to memorize and used it as a form of steganography.