Memory constraints in ye olden days meant that text-mode display adapters had room for either upper or lower case, but not both. Why was this universally uppercase and never lowercase?

I remember reading that some programmer felt that if someone needed to type "God", it wouldn't be respectful to have His name in lowercase, so he went with uppercase. I don't think my fellow Christians would be too bothered by this, so it seems more likely to me that it's just a remnant of the old telegraph system. But does anyone really know why?

  • 30
    I wonder how much impact the fact that the lettes j, g, p, q, and y dangles below the baseline had. In an English typeface the capital letters doesn't have anything below the baseline.
    – UncleBod
    Commented Dec 14, 2023 at 7:26
  • 34
    Historically uppercase was first. Ancient Latin and Greek only used capital letters. Lowercase are a late invention to make hand writing easier and are a derivation of the original graphs. Commented Dec 14, 2023 at 7:38
  • 17
    When I was young I played a lot of really loud music. Using my computer was easier when it was shouting at me. :-)
    – Geo...
    Commented Dec 14, 2023 at 14:12
  • Comments have been moved to chat, which you should totally check out. Please continue the discussion there.
    – wizzwizz4
    Commented Dec 16, 2023 at 20:38

7 Answers 7


TL;DR: Upper Case Is the Default for Latin Script.

Latin script is based on upper case and designed around that. Lower case is a later add-on (see below) for cursive. Default use-case for Latin script is Upper-Case.

Character Encoding

Early codes and machinery used for writing didn't have any case at all, only letters. Lower and upper case glyphs are eye candy, usually not conveying any substantial meaning. That's why neither Morse nor Braille carry them. Same goes for any other early transmission code, like the the ITA2 used for teletypes.

After all, there is no need to have two sets when they are fully complementary. It wasn't until the mid 1960s that, with the upcoming ASCII encoding, lower case became a choice to be used.

Printing Requires a Choice

The whole point of choosing what case to be used only arises with the need for displaying those letters in print, or later on screen. Here using upper case is only natural, as

Upper Case is

  • most commonly accepted

    across Europe, independent of culture and writing style due to being the oldest form.

  • most readable,

    due to being a two line script (*1) developed for best readability (*2) - a feature that comes in quite handy with low quality output (*3).

  • most inclusive / least offensive one.

    Writing all all upper case will always be proper, while writing all lower may be offending to some - such as when addressing a person.

Then There Was Data

When it comes to data handling, it's important to remember that letters were only added as an afterthought. Punch cards, the earliest means of data storage, only had numbers at first. Next, some punctuation (decimal point, currency sign, hash, etc) were added, and finally letters were the last major modification.

The same goes for equipment to print from punch cards (printing tabulators) or onto punch cards (interpreters). They originally also could do only numbers and punctuation: letters were added later. Powers first in 1921, while IBM followed in 1931. Of course, when printed they were done so in upper case.

It wasn't until the mid 1960s that lower case became a distinct, additional set on punch cards and mainframes.

But I Type Lower Case

Not really. Early machinery, no matter whether teletype or terminal, did not produce lower or upper case when typed, but just letters, as there was no case (see above). This single case was usually displayed as uppercase (reasoning see above).

The Apple II is eventually the last major application of that principle. pressing a letter always produced only a single code, no matter whether Shift was pressed or not, and this letter was displayed in upper case.

Almost all early computing equipment worked similarly. When terminals allowing entry of lower and upper case letters became a thing, mainframes still only got a single case delivered as communication equipment (terminal controllers, I/O handler) simply converted both to the single encoding the host used.

That way, terminals with lower case and those without would by default be operated all equally, allowing a smooth transition and high degree of compatibility.

Then Why is Unix Lower Case?

The main purpose of low end machines of the 1970s was about to get any work done. Translation layers, such as the ones common in mainframes, were a luxury one could do without. Unix was a prime example as it was all about getting a certain functionality to production. Combine this with now widely available (ASCII-based) lowercase enabled terminals and one gets lower case as default input. Including all the hassles of case sensitive input.

CP/M and in turn MS-DOS are nice counter examples as here the command line is by default translated to upper case, simplifying handling.

Bottom Line

There is no single reason and especially not one constructed in afterthought. It was an evolutionary process influenced by need, capabilities and usage.

History is Upper Case

  • Latin, developed around 7th century BC was originally a two line script(*1).

  • By the first century AD papyri show sloppy written letters forming lower case (minuscule), which developed into Roman Cursive.

  • By the 8th century those forms had developed into Carolingian Minuscule similar to today's lower case (*4).

  • It took another 300-500 years (11th/13th century) for Gothic script (Blackletter) to finalize them into what we know today.

  • Rules of when to use upper and lower case only became widely accepted during the 16th/17th century, not least by development of the printing press.

  • The English way of writing mostly lower case is even more recent and dates to the early 18th century.

Bottom Line: Throughout history it was Upper-Case first.

*1 - Two line scripts, like old Greek and Latin, use a single case with all letters of the same height, limited between those two lines. Lower case needs three lines, or four with descender (itself a later, Gothic addition). Might remind you about school, perhaps? :))

*2 - Roman Majuscule developed to be readable from distance and on low quality material. Roman Minuscule is a form of shorthand, trading readability for space usage and speed in writing(*5). More appropriate for notes and private correspondence than official exchange.

*3 - Note that your keyboard is (most likely) also labelled with upper case letters. That's a convention since early typewriter days. They are until today more easy to identify than lower case - so not (only) heritage.

*4 - Greek Minuscule was developed in Byzantine Empire about 100-200 years later, which also explains why Cyrillic has distinct lower case forms for only some of the letters as it got defined before the process of developing lower case for Greek was concluded.

*5 - Ligatures make a great point here. They were a further evolution promoting increasing density and speed of cursive - and almost exclusively found with lower case letters. While usually not considered letters on their own, typesetters had dedicated letters for each ligature.

  • Comments have been moved to chat; please do not continue the discussion here. Before posting a comment below this one, please review the purposes of comments. Comments that do not request clarification or suggest improvements usually belong as an answer, on Retrocomputing Meta, or in Retrocomputing Chat. Comments continuing discussion may be removed.
    – wizzwizz4
    Commented Dec 15, 2023 at 15:44
  • Why are you empathising Latin script, if in Cyrillic and Greek it is the same case?
    – Anixx
    Commented Dec 19, 2023 at 7:41
  • @Anixx Because Latin is what early computers/OS/processing equipment used. Other scripts are later, country specific add-on. Also, Latin case predates Greek any Cyrillic (see footnote 4)
    – Raffzahn
    Commented Dec 19, 2023 at 10:01
  • @Raffzahn if we consider telegraph, it used caps-only for cyrillic as well, of course, it appeared before computers. Since it used 5-bit codes, it did not have characters even for dot, not to say, Latin. ar.culture.ru/attachments/attachment/preview/…
    – Anixx
    Commented Dec 19, 2023 at 10:05
  • @Anixx What's your point? ll of that is already in the answer. And no, Baudot and later ITA-2 is fixed assigned to Latin letters (including dot being 11100) since the beginning. Since Cyrillic letters are not the same as Latin,it's not just an exchange of glyphs. MTA-2 was only standardased as alter as 1963. (Also, note, Russian telegraphy started with German equipment :))
    – Raffzahn
    Commented Dec 19, 2023 at 10:33

I would venture to say that it stems primarily from the telegraph tradition, starting with Baudot (1870s) which eventually led to ITA2. Essentially 5 and 6 bit codes (arguably a 5-bit code with a shift code used frequently is nearly a 6-bit code) based on upper case + digits + a handful of formatting (carriage return, line feed, space) and special characters. Teletype equipment was used for I/O on many early computers. The Teletype 33 was an evolution of this equipment, incorporating the ASCII character set except for lower case characters - purpose built for use with computers but still UPPER CASE only.

Lower case, or more properly mixed case, text improves readability. But it doesn't, generally speaking, change the meaning of text. For a telegram, news report, computer program, utility bill, stock market report, etc. - UPPER CASE will get the message across just as accurately and effectively as mixed case. When each additional character requires additional mechanical parts to produce it, increasing the character set by 50% just to add lower case characters is an expensive proposition. The first commercial dot-matrix printers arrived in 1968. Prior to that, most small printers were based on teletype and similar equipment with a separate mechanism of some sort for each character. One notable exception is the IBM Selectric, and even that did not come on the market until 1961.

All of this wasn't "display adapters" or video terminals. This was electro-mechanical hardware. It makes sense that the first generation of video terminals copied these constraints, though with the restriction being the cost of memory rather than the cost of mechanical parts. But always UPPER CASE because that's the heritage from the 1800s.

  • 3
    I think it goes ack even further ack to antiquity. Latin and Greek only used uppercase letters. The writing system is based on uppercase. Lowercase is just a derivative of uppercase, not the otherway round. Commented Dec 14, 2023 at 7:41
  • 9
    @PatrickSchlüter /Surviving/ Latin is uppercase, because it was monumental. On a day-by-day basis miniscule was used on occasion, but relatively little has survived due to the impermanence of the media en.wikipedia.org/wiki/Majuscule#History Commented Dec 14, 2023 at 13:50
  • 1
    I guess that just pushes the question back to why teletype used upper case. Commented Dec 14, 2023 at 16:33
  • 1
    @MarkMorganLloyd: Even today, many people's handwriting is somewhat different when writing things for other people to read, versus scribbling temporary notes for their own use, and while today people use paper for almost everything, historically a lot of temporary note-taking would have been done on was tablets which could be easily reused, but would be unlikely to survive for thousands of years without deliberate efforts at preservation.
    – supercat
    Commented Dec 14, 2023 at 17:05
  • 1
    @gidds The again there are many more even without looking at case - and spoken language is even more ambiguous. After all, meaning is never absolute but transferred in context.
    – Raffzahn
    Commented Dec 14, 2023 at 21:35

Teleprinters were in widespread use for decades before computers arrived. Teleprinters (Teletypes, etc) could read and write text on punched paper tape. They were the first interactive i/o adapted to computers. The teleprinter code was 5 bits, and the tape was 5 holes wide. 5 bits was enough for upper case letters, numbers, a few symbols, and control characters. Computers were already in their 2nd generation when teleprinters using 7-bit codes with upper and lower case became available. https://en.wikipedia.org/wiki/Teleprinter Interestingly, the teleprinter code fits 26 letters (A-Z), 26 figures (numbers and symbols), and 6 control symbols into a 5-bit code. (5-bits represents exactly 32 symbols). The trick is that the "letters" and "figures" are two character sets. The choice of active character set is modal and is changed by the code elements LETTERS (11111) and FIGURES (11011). Teleprinter code elements (image from wikipedia)

  • 3
    5 bits is not enough for upper case + numbers etc. - 5 bits is only 32. It really took a bit of a hack (shifting) to get all that into 5 bits, arguably making it a nearly-6-bit code. And 7-bit codes - e.g., ASCII - didn't necessarily result in lower case being distinguished from upper case - see TTY 33 for the classic example. Commented Dec 14, 2023 at 15:03
  • It's an interesting experience to work with one of these for those of us that were born after these things were dying off. A few historic computing museums around the country have functioning teletypes. Less and less people realize there was a time before programming in a cloud ide from an apple watch.
    – Rig
    Commented Dec 14, 2023 at 17:48
  • 5 bits would also have been enough for lower case letters, numbers, a few symbols, and control characters. So, while "because teleprinters used upper case" might be a sufficient answer for why computers used upper case, it passes over why upper case was chosen over lower.
    – dave
    Commented Dec 15, 2023 at 3:08
  • The question then becomes "why did teleprinters use upper case?" The answer to that can only come from the individuals who designed the machines in the late 19th and early 20th centuries. Since they and their reasons are long gone, we might look to the field of folk typography to aid in our speculations.
    – Ed_B
    Commented Dec 15, 2023 at 6:23
  • 2
    Important to remember that, until computers began using monitors, they didn’t even determine which letter shapes were displayed. Their terminals did.
    – Davislor
    Commented Dec 15, 2023 at 8:30

Uppercase letters are more distinctive.

Additionally, you typically learn uppercase letters first, as they are considered the easiest, followed by lowercase letters, and then more complex handwritten forms.

Also the telegraphy tradition surely contributes as well (and it probabely was led by similar reasons).

One small observation - my keyboard (and any I can remember) have uppercase letters printed on keys as primary markings even now :)

  • 2
    This is particularly relevant for low-resolution displays, like analog TVs
    – Chris Dodd
    Commented Dec 14, 2023 at 6:40
  • 3
    And even more for teletypes, printers and such, which was the output medium at the time, while displays came in much later. And especially as the quality of the prints degraded as the printers was used and the printing tapes was exhaused and so on, as life of tools goes.
    – gilhad
    Commented Dec 14, 2023 at 6:49
  • 1
    I have downvoted this answer as I believe it is wrong. In my opinion the correct answer is more about what input/output devices where available at the time the computers where made and the languages where defined. A typical device used was the ASR33 (introduced 1963) that only had upper case letters.
    – ghellquist
    Commented Dec 14, 2023 at 13:27
  • 2
    1963 is quite late, though. Computers were printing things out before the ASR33.
    – dave
    Commented Dec 14, 2023 at 13:33
  • 1
    @ghellquist the preference of upper case in general and for information processing in particular predates the ASR33, thus it can not be of any reason. More important, citig asingle device des ot answer the question, it pushes the can only down the road (as in 'why did the ASR33 use upper case').
    – Raffzahn
    Commented Dec 14, 2023 at 13:34

In addition to the answers already given: technical limitations.

Lowercase characters could often not be displayed or printed properly. A nice example is the HP-41C, which had a 14-segment display and could easily display uppercase characters, but had troible with lowercase characters (it did support the characters a..e).

Matrix printers (modern back n the days!) had 7 pins, allowing for lowercase characters...but needed to shift characters like g,j,y upward which was, let's say, inelegant.

Lowercase characters simply needed more space to be displayed properly, as they extended below the baseline. And being otherwise often more "complex" (tighter loops, more shorter lines being closer together) than their uppercase counterparts, they were harder to read at low resolutions.

  • 2
    But it goes much farther back than matrix printers. The early computer builders adapted the tech in hand, which was telecom and tabulating (punched card) gear. That used upper case.
    – John Doty
    Commented Dec 16, 2023 at 14:41
  • 1
    @John Doty Of course you're right. But the original question was about "older" computers, not "the oldest". The mentioned HP-41C came out in 1979, so it's not exactly "ancient", but it's one case where the technical limitation is vey obvious. And yes, it was referred to as a "personal pocket computer" back in the days.
    – Klaws
    Commented Dec 18, 2023 at 8:44

I would assume the main reason was the availability of input/output devices. Mechanical printers and teletypes are getting increasingly more complex the more characters they need to support - going with the minimal set of characters was thus reducing complexity. Most (with some exceptions) systems were using 7- or 8-bit character sets anyways, so, in theory, you could have used much larger character sets once deemed necessary, but the limits were in the mechanical devices.

That leaves us with the choice of upper case vs. lower case. Here, all upper case is simply easier to read than all lower case, because characters are more distinctive (because they simply use more "paper area"). I pretty much doubt your anecdotal quote regarding "god", though. This word was pretty much not expected to occur very often....

  • 2
    "Here, all upper case is simply easier to read than all lower case": I beg to differ on that; it's the whole reason lower-case was developed at all, to be easier to read/write. Commented Dec 14, 2023 at 12:43
  • 3
    @OmarandLorraine Write (by hand) I perfectly agree. Read, I don't.
    – tofro
    Commented Dec 14, 2023 at 13:03
  • @OmarandLorraine lower case was developed as shorthand to speed up writing (and saving space on expensive paper), not reading. It contains way less features for distinction than upper case. After all, reducing features is a base process of compression, isn't it?
    – Raffzahn
    Commented Dec 14, 2023 at 13:24
  • 2
    @tofro -- The second answer that gets is completely wrong, sorry. Uppercase is less distinctive, distinctly harder to read, this really is typography 101, the very first thing anybody starting to work in press learns. This is the reason why typesetting never used uppercase for larger runs of text.
    – Gábor
    Commented Dec 14, 2023 at 20:32
  • 2
    For example, if you have a cage that has the word of the three-letter animal written on it, cat, bat, dog, pig, fox, cow, ape and elk can all be distinguished just by the pattern of the ascenders and descenders. In capital letters, CAT, BAT, DOG, PIG, FOX, etc. all have same general shape, a rectangle.
    – prosfilaes
    Commented Dec 14, 2023 at 22:23

Before the availability of affordable video display adapters, the original language definitions for BASIC and FORTRAN were created containing upper case keywords. So the earliest text-mode (e.g. not just switches and blinking lights) personal computers which had limited character ROM space for character sets, included the character set that allowed these video-based computers to be programmed in BASIC.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .