Some alphabets, e.g. the Greek, Hebrew and Arabic alphabets, have different forms for some letters when they appear at the end of the word.

E.g. in Greek, the letter sigma (σ) appears as a ς when used at the end of a word; in Hebrew, a couple of letters have a special final form; for example the 'm': מ becomes ם when used at the end of a word.

While separate forms for capital letters has a definite advantage (marking the begin of a sentence, or indicating proper nouns) which increases the readability, I can't see any advantage of having a separate final form for certain letters. In fact, this makes an alphabet harder to learn for children/foreigners.

Are there any practical advantages to having final forms for certain letters, or is it just a 'stylistic' choice which survived throughout the ages (unlike the 'long s' in the Latin alphabet)?

  • 2
    I believe the majority of such cases are in alphabets deriving from the Syriac alphabet, and it's probably a natural result of stopping writing in a cursive alphabet. – user6726 Oct 16 '16 at 1:05
  • @user6726. Greek script derives from Phoenician, Hebrew square script from Old Aramaic, Arabic from Nabataean. None of these derives from Syriac script. – fdb Oct 16 '16 at 9:39
  • True: I was speaking of the numerous other scrupts such as Mongol, Manchu, etc. where final forms are endemic. – user6726 Oct 16 '16 at 13:36
  • @user6726. Mongolian script derives from Sogdian, not Syriac script. – fdb Oct 16 '16 at 20:48
  • 1
    @user6726. Imperial Aramaic. – fdb Oct 17 '16 at 8:28

Usually the final forms weren't designed intentionally. They arose over time through, effectively, sloppy handwriting.

Up through the mid-Hellenistic period, sigma's various forms (from the same root glyph as "Ш", "ש", and "ش") evolved into standard Σ. This shape was used in inscriptions and important writings. But scribbling Σ over and over gets tiring, and the left side needed to be relatively flat to avoid confusion with Ε. So in quick, informal handwriting, it ended up looking more like a Ϲ, the "lunate sigma".

In cursive script, then, the shape would be connected to the letters on either side. If there was a letter immediately after it, the Ϲ was rounded off into a circle, with the top emphasized to keep it distinct from ο: thus, σ. If there was no letter immediately following it, the line just trailed off: ς.

Then during the Renaissance, printers typesetting Greek text began imitating the conventions that had arisen for other languages: using uppercase and lowercase letters to break up words and sentences and make reading easier. For uppercase they used the classical inscription letters, and for lowercase they tried to imitate cursive handwriting.

This gave us the three modern forms of sigma, Σ/σ/ς. There was no real reason to distinguish σ from ς any more, but at that point it was what readers were used to. Much like capitalizing the pronoun "I" in English, the reasons for it are long gone, but tradition and inertia keep it in place.

EDIT: My information about Hebrew was incorrect, so I won't weigh in on that case.

  • 1
    What you say about Hebrew is totally wrong. The medial forms are the older forms. Also: In Hebrew you never "connect (a letter) to the next letter". Each letter is written separately. – fdb Oct 15 '16 at 22:13
  • "sloppy handwriting?" I don't think so. – mobileink Oct 15 '16 at 22:39
  • @mobileink I was being somewhat facetious, but it's a process of accumulated "errors"/changes. Like how Italian descended from what was at the time considered incorrect Latin. – Draconis Oct 16 '16 at 0:07
  • @fdb It seems I was misinformed. Thanks for pointing that out! I removed that section from my answer until I can find a better source. – Draconis Oct 16 '16 at 0:08

It makes total sense to signal the start/end of a word in a ScriptThatDoesNotUseWhitespace. The ancient scripts that I have seen do not use whitespace. Having individual initial forms like capitals next to terminal forms as in Arabic

In a script that does have whitespace, terminal forms may come pretty natural if writing cursive, that is a handscript which does preferably not lift the pen until the end of the word--in contrast to American hand lettering, which mimics print letters. I totally agree with Draconis on that part, although that might look as simple as simple curl or drawn out stroke, as a matter of caligraphy. The economics of handwriting do have to be taken into account. There is a big difference between imprints, scratches, or brush strokes. This explains miniscule "e" as a capital "E" starting with the middle stroke but connecting to the above stroke. Whereas a stonemason can work more effectively if avoiding curves and extraneous strokes.

Another thing to consider is that terminal phonemes do possibly have a different sound. It's possible to see a kind of graph-symbolism in letters that imitates sound-qualities. I am not sure what to call this, since sound-symbolism means the exact opposite. Just for example: "I" is thin, "O" is wide and round like the rounded lips, "U" is open, "A" like "K" shows an increasing envelope, what with the glottal plosive that it often is, while "E" or rather Greek "H" (Eta) is very centric with it's central stroke.

The display of the letters depends on the font you are using, but I have no time to take pictures or draw animations.

i do not know enough about Arabic or Hebrew caligraphy for example, so I cannot speak to that.


Arabic has no fewer than 4 forms: independent, initial, medial, and final. But in all cases there is a "kernel" form common among them all. One advantage of this design is that you do not need spaces to separate words – not a minor consideration if writing materials are expensive.

Caveat: in Arabic, final forms are not always distinct. For example, a final Waw can look just like a medial, initial or independent Waw. But the system works.

  • waw has only two forms: independent و and final (right-joined) ـو. It does not have an initial (left-joined) and medial (left-and-right-joined) forms. – fdb Oct 16 '16 at 9:21
  • 1
    @fbd: very good. you wiil notice that i did not say that every letter has four forms. i hope you can see that "Arabic has 4 forms" does not entail "everything in Arabic has 4 forms." in any case that has littlevrelevance to the op's question. and as i noted, the advantages of having final forms is efficiency. this is very clearly evident in manuscripts. – mobileink Feb 15 '17 at 21:31
  • …surely Arabic does have spaces to separate words, which is why final forms exist? – Draconis Jul 5 '19 at 4:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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