Please send more stamps

A little toot on the Isn't Bobby Clever horn?

I don't do Fiendface or Instah, but over the last several years, I've tuned into Metafilter for my para-social life. At the end of last month, Paul Slade, a British MeFite asked a question about some WWI post-cards which he'd bought. We've met Slade before with his long-form investigation of Kit Williams and the Masquerade hare.

It seems that a clutch of post-cards had been written home in 1915 by a future Tory grandee while he was in boarding school at the other end of the country. It was relatively easy to work out who Robin the Author was but one of the post-cards was covered with numbers and symbols whose meaning was harder to make out than the regular upper-class intra-family slang which filled the other missives. 

Eeee but I do like a good puzzle! I've clipped part of the message [picture top] in the scraggy hand-writing of a 12 y.o. for whom penmanship was not on the curriculum of his very expensive education. When my mother died in 2020, I inherited a small plastic bag containing my own letters home written almost exactly 50 years after young Robin's. My handwriting was terrible too. It wasn't until I started inventing myself aged ~15-16 that I developed a reasonably consistent cursive hand. By then I'd twigged that good handwriting is quicker than scrawl, quite apart from being easier to read.

It's not much to go on (30 words = 150 characters) but if you've inhabited the mind of a small chap writing home from  an Institution, it's defo easier to imagine what could be written: it's rather a reduced instruction set of concepts. Have a go, yourself? But I'll not hold it against you if it's couldn't be arsed. One approach:

• Vowels 
• My first observation was that the message is a mix of numbers, letters and symbols; but that only the numbers 1-5 occurred. I hypothesized that these might represent the 5 vowels a e i o u in that order. 
• 3 appears twice as a single word, so that has to be a or I, the latter being more likely. 
• 2 is far more common than the other digits which accords with the high frequency of e in normal English. 
• English allows double vowels ee and oo, but not aa ii uu so the 44 in line 6 is likely oo. 
• The diphthongs in the message are also consistent with my hypothesis in standard English 45 = ou;  21 = ea; 32 = ie. I'd have rethink my program if the text was indicating ua ia or  ui
  

• Consonants 
• with vowels sorted, similar thinking can be used to assign each of the symbols to English consonants. 
• The first line starts dea? which gives us the symbol for r. 
• Doubles: The last word in line 3 is ?eKK. There aren't many possibilities there ?ebb, ?ecc, ?edd aren't in common English but ?ell is, with ?ess also possible. 
• If that word does end in ll the next word is ?ould which must be could or would. That l also helps with the last word on line 4 ?lea?e. Crossword solver [or similar] is your friend: that combo must be cleave, oleate, please, or sleaze. The same word appears on line 4 but also in the PS at the top of the card - gotta be please. Young Robin being a politely demanding chick still. This sub-solution is handy because s is the 3rd commonest consonant after t and n and appears 10x in the message.
  
• More doubles 
• M5MM? appears twice at the end of the message - could that be Mummy?
• From what we've already deduced, the first word in the last line reads le??er, which could be lesser or, if Robin is Afrikaans, lekker but I'll bet on letter. 

And there you have it. I think the standard operating procedure outlined above would have utility for any simple substitution cipher. I wasn't the only person to solve the puzzle. Answer on Metafilter.