What's next

I recall seeing an article from 1969 about musical chord recognition: the psychologist was diligent and sought to relate her theory to neural architecture of the inner ear. I don't know if she was successful but I do know I'm not even trying. I cannot say or even imagine what the brain-cell embodiments of moments or memories, loud or soft, booming or peeping, are. That's the chief and obvious shortcoming of this model.

But that's the chief and obvious shortcoming of most psychological models. They can still propel useful inquiry. This model suggests (if you think about it!) the following:

• If you can increase "volume" of present and future moments, that will help. However, if "reverb" is at all high, for either past or present moments, that creates problems. Note the model simply postulates that "reverb" drops 1 nameless unit at a time, each time a new moment presents itself, but all that is just an invention.
• Therapeutics: reduce reverb all around, or past volume. But the model defines those as immutable. Well, old stuff is, once it's in memory. If new stuff's reverb and/or volume can be "set" the first time you experience it, that may help later. "Therapy" may not be immediately possible but "prevention" might amount to it later.
• Preventives: get MORE memories? If it is inevitable that the present get softer, can it be hoped that a crowd of relatively recent soft ones will conceal or obscure the old loud one? The loop as I have programmed it stops at the first match it finds. If that is really how your head works, then repeating or duplicating old moments might be useful. And also - somehow - keep their reverb as well as their volume low. I wonder if there might be an exercise to LIMIT reverb upon processing a freshly received moment. That is, whatever your mind's habits are these days, forcibly change them. Perhaps SPEED of incoming moments can do that. If you voluntarily linger over a moment, does that MAKE its reverb high? Or is that irrelevant, and it is really the content of a moment - in this case, the length of a proposed itinerary - that determines how long you will tarry?
  
  The model as it stands makes no provision for any of this. What is called in the JavaScript blankDrawer, the proposed itinerary for which there is no memory, is not itself stored, but maybe it should be. And maybe there should be several of them, with non-uniform lists of states. How soon you click Tick-tock or Airball after one, and/or how many states are named in it, could in later models determine its reverb.
  
  
• Diagnostics: how would you MEASURE or at least infer reverb and volume, with a view to assessing changes in either? Invocation of a long-ago unique memory might be practical. Such a memory might be the license plate on a car you owned long ago: once you got rid of it, there would be nothing to reinforce that memory. Anyway, if revisiting a certain memory NOW gets in the way of revisiting another, that is significant. So, need 2 diagnostic memories? How about that old license plate and your current one?
  
  But note these are semantic memories, not episodic ones. By the way, does revisiting a memory make a new memory? And if that new memory has to be loud, does that make things worse for further revisitations? Maybe not, if the revisited episodic memory is itself a semantic one - which in this model is assumed to be always and instantly recoverable. I should admit, though, that that is just negligence on my part: it was easy to program nothing at all about such memories. In any case, if their durability and accessibility are real, then regardless of the model, a great therapeutic/preventive would be to convert old episodics to new semantics. Hmm.
  
  

Though it is not necessarily useful for computer modeling of dementia in general, it is useful here to keep in mind what is being modeled here: not loss of memory, but incontinence of memory. That is, remembering not too little but too much, and then being fuddled by all those memories, to the point of forgetting (so to speak) that they reside in the past - they are not the present, as much as they may feel that way. That the past is always there, and is always fully recoverable, this model postulates. In early 2013, in Tabatinga, I sat at an open-air bar and a middle-aged patron who was so drunkenly accommodating I took him for an employee reminded me, for the first time, of another Brazilian bar experience I'd had, way back in 1989. That one was in Chapada dos Guimarães, and may have been overshadowed by activities the next week; anyway, I remembered it at last, the genial drinking, the sudden silence, and the stranger turning to me and asking, "Say...weren't you in Japan last year?" After an amazed interval, I answered: "No." Bringing the house down, he said, "Well, neither was I!" It was quite gratifying to see this mini-movie all over again. But viewing it, I never forgot I was still in Tabatinga. Well, Tabatinga was louder!

Back to the model's details
Back to the model itself
Home