Alphabetic Design Elements - or: I've had too much coffee!

  • Thread starter Thread starter DaveC426913
  • Start date Start date
AI Thread Summary
The discussion revolves around the creation of a unique alphabet design using a maximum of 14 defined elements, inspired by a late-night coffee-fueled brainstorming session. The designer chose Century Gothic for its simplicity and faced challenges with letters like v, x, and z due to their overlapping features, necessitating additional symmetry elements. The analysis reveals that while some letters require fewer elements, others like 'g' are over-specified, leading to inefficiencies in the design process. The theoretical minimum of element types needed to uniquely define all 26 letters is suggested to be around 4.7, prompting further exploration into the potential for fractional element types and sub-elements. The conversation highlights the complexity and creativity involved in alphabet design.
DaveC426913
Gold Member
Messages
23,829
Reaction score
7,814
This is what happens when I have coffee after 2PM. (I am not a coffee-drinker.) Did this in my head between 12 and 1AM while staring at the ceiling of my bedroom.

I've uniquely defined the letters of the alphabet with a maximum of 14 design elements.

1703089924289.png

I chose Century Gothic as a typeface with the simplest, most basic, elemental styles (sans serif and no funky tails on 'a's). I cheated with 'u'. It doesn't actually have a spine in CG, but, in my world 'u' is best defined having a spine.

I had the biggest struggle with v,x and z. Element-wise, they have a lot of overlap. I finally had to add two "Symmetry" Elements just for these cases, (which, as you will see below, is a very inefficient use of parameters).

Some are sort of default characters. 'c' requires only one element type.

Some are on the long tail of the frequency curve, with 'g' having the most elements (5). Now, that is a little over-specified - g is already uniquely-defined without needing to specify that it is counter-facing. But since I had to add 'Counter-facing' to distinguish p from d and q from b, it just incidentally applies to g. Similar thing happens for horizontal symmetry with 'i' and 'l'.
 
Last edited:
  • Like
Likes Charles Link
Physics news on Phys.org
1AM till 2AM14 element types to define 26 unique characters is not great, really. It takes .54 element-types per character, which is really wasteful.If we take a side-trip into abstracted optimization, the theoretical minimum element-types is 5. 2^5 = 32 which is enough in theory to uniqiuely define every letter:
1703090145439.png


(though they wouldn't appear in order like a binary sequence graph):
1703091133743.png
This solution is highly theoretical because it is going to be very difficult to find a mere 5 design elements that, in combination, uniquely define each of the 26 letters.

Glossary: Element
An element type has a loose set of conditions:
  • It must be a common element (i.e. 2 or more characters require it; so "x-shaped" doesn't qualify as a element type)
  • It must be a simple, basic construction element ('diagonal' is a basic element; 'shaped like a p' applies to both p and b but doesnt qualify as simple/basic).
5 elements to define 26 characters is pretty good. That's .19 element-types per character.

But even 2^5 is a little high for the theoretical minimum. (As long as we have the ability to define 32 characters uniquely, we will always wonder: what do those extra 6 possible characters look like?**)Technically, we really only need 4.7 element types (2^4.7 = 26). That gives .18 - the best possible ratio of elements to characters.

Is it possible to have a fractional element type? I think it is. Some elements are subsets of others, so they don't have complete independent range.

Ascender/descender are mutually exclusive (well, not by rule, but by there being no such character). It might be possible to formulate a sub-element type. And there's a lot of overlap between 'tail' 'arch' and 'open loop'.

More work to be done...**

1703091582609.png

https://en.wikipedia.org/wiki/On_Beyond_Zebra!
(definitely more than six!)
1703091757974.png
 
Last edited:
Look at the way Chinese characters are entered on a western keyboard.
https://en.wikipedia.org/wiki/Wubi_method

In August 1983, exactly 40 years ago, a Chinese engineer named Wang Yongmin developed the first popular way to input Chinese characters into a computer: Wubi. He did it by breaking down a Chinese character into different strokes and assigning several strokes to each letter on the QWERTY keyboard.
https://www.technologyreview.com/2023/08/23/1078274/fascinating-evolution-typing-chinese-characters/
 
  • Like
Likes DaveC426913
Similar to the 2024 thread, here I start the 2025 thread. As always it is getting increasingly difficult to predict, so I will make a list based on other article predictions. You can also leave your prediction here. Here are the predictions of 2024 that did not make it: Peter Shor, David Deutsch and all the rest of the quantum computing community (various sources) Pablo Jarrillo Herrero, Allan McDonald and Rafi Bistritzer for magic angle in twisted graphene (various sources) Christoph...
Thread 'My experience as a hostage'
I believe it was the summer of 2001 that I made a trip to Peru for my work. I was a private contractor doing automation engineering and programming for various companies, including Frito Lay. Frito had purchased a snack food plant near Lima, Peru, and sent me down to oversee the upgrades to the systems and the startup. Peru was still suffering the ills of a recent civil war and I knew it was dicey, but the money was too good to pass up. It was a long trip to Lima; about 14 hours of airtime...

Similar threads

Replies
5
Views
2K
Replies
7
Views
3K
Replies
1
Views
3K
Replies
2
Views
9K
Replies
1
Views
3K
Replies
1
Views
3K
Replies
5
Views
4K
Back
Top