Unicode has the following APL characters that don't seem to be used in modern APL. What functions or operators did they represent? Are they still supported for backwards compatibility, but undocumented? I have also seen some of these in code that's no longer in use.

  1. Quad circle
  2. Del tilde
  3. Quad jot
  4. Circle jot
  5. Quad slash , also quad backslash
  6. Leftwards vane , also
  7. Quad delta , also
  8. Delta underbar
  9. Del diaeresis aka umlaut
  10. Quad question
  11. Tack underbar
  12. "Down shoe stile" , looks like Greek psi

Also, has Infinity, , ever been supported by an APL?


2 Answers 2


What functions or operators did they represent?

Adding to LdBeth's answer:

  1. is used to lock functions in almost all mainstream APLs (Dyalog APL is the exception), and used as the obverse operator in dzaima/APL.
  2. is on IBM APL2's language bar, but as far as I can tell, it isn't in use.
  3. is used in dzaima/APL as table (harmonised syntax for ∘.).
  4. is used in dzaima/APL as oblique. is on the APL2 language bar, but as far as I can tell, it isn't in use.
  5. was used by APL/700 to create, rename and change password for files, and was for deleting files.
  6. is a perfectly normal identifier character in most APLs.
  7. hasn't been used yet, but is proposed for lazy evaluation in Dyalog APL.

Are they still supported for backwards compatibility, but undocumented?

One could argue that this is the case for .


APL used to be typed by a typewriter in its early stage.

Say if you want to type , you first type =, then press backspace, then overstrike with /. (The order doesn't really matter, it is also ok to type / first).

The typewriter is connected with the mainframe that is able to figure out the "composition" of different characters.

Many of these combinations are never assigned a meaning, but a fews do have been use by some implementations or at least has proposed meaning. Unicode includes these to served as "future extensions" to APL language, instead of for compatibility reason (well it could be said these are for compatibility with old terminal devices).

I can recognize some of the listed glyphs:

is assigned as Under, listed in https://dl.acm.org/doi/abs/10.1145/3386319

f⍢g ⍵ ←→ (g⍣¯1) f g ⍵
⍺ f⍢g ⍵ ←→ (g⍣¯1) (g ⍺) f (g ⍵)

works like as Del editor, but would create "locked" function that end user is unable to modify it. from APLX. btw APLX also uses ,,, for file IO.

, , , , are used by NARS2000, you can consult the documentation of it. http://www.nars2000.org

  • 2
    Amazing. It's a wonder that APL isn't more popular.
    – JimmyJames
    Dec 7, 2021 at 16:17
  • 2
    @JimmyJames - I upvoted your comment because it speaks a truth - and I like APL! (And frankly, I like APL's special character syntax much more than derivative ASCII syntaxes like J and other follow-ons, that stuff just looks like line noise.)
    – davidbak
    Dec 7, 2021 at 20:49
  • 1
    I can see how it would appeal to someone who feels at home with mathematical symbols and formulas. That kind of thing was always a barrier for me: math was far more accessible when explained in words. But if you like it, cool. To each their own. It reminds me of hieroglyphics.
    – JimmyJames
    Dec 7, 2021 at 21:03
  • @JimmyJames, "It's a wonder that APL isn't more popular" - apparently it was once fairly popular, though I suspect Excel today appropriates most of its would-be users. I think the fundamental error was that it's syntax and symbology was over-trained both to suit the particulars of 1960s IBM technology (including things like overstruck characters that were later difficult to reproduce on PC technology), and to suit the preferences or workloads of mathematicians as opposed to business programmers. Iverson later saw the mistake, but by then other more English-like languages had overtaken.
    – Steve
    Dec 8, 2021 at 9:12
  • @Steve I'd wager that the programmer base at the time APL was relatively popular was both much smaller and also much more heavily weighted towards scientists and mathematicians. I've recently learned that the notion of pure symbolic human languages is a myth, they all have phonetic origins. APL seems to fit that concept at some level at least and I do find that interesting. Maybe I'll try to learn more about it.
    – JimmyJames
    Dec 8, 2021 at 14:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.