The standard placeholder text is a corrupted passage from Cicero, written in 45 BC, arguing about the nature of pleasure and pain. The first word is not a word: "Lorem" is what remains of "dolorem" after the start of the sentence was lost.
The passage says roughly that nobody loves pain for its own sake — which is a small joke on anyone who has spent an afternoon fighting a layout.
Why has it lasted?
Because it is word-shaped and unreadable at the same time. A reviewer looking at a layout filled with real English starts reading it and stops looking at the design; a layout filled with Latin gives the eye something with the right texture and nothing to follow.
It survived the transition from print to screen twice. It spread through dry-transfer lettering sheets in the 1960s, then through desktop publishing software in the 1980s, and by then it was the default that every subsequent tool copied.
The corruption is part of the value. Recognisable Latin would invite translation; a mangled version reads as pure texture.
What does it hide?
The content problems. Placeholder text is uniformly medium-length, uniformly punctuated and never awkward, which means a layout designed against it has never been tested against a real headline that runs to two lines or a product name with no vowels.
It also has the wrong statistics. Latin uses fewer short function words than English, so line breaks fall differently and the ragged edge of a paragraph looks different — a column that looks balanced in Latin can look untidy once the real words arrive.
And occasionally it ships. Placeholder text appearing on a live page is common enough to be a running joke, and it happens because it looks deliberate enough not to catch the eye during a final check.
What is the better practice?
Real content where it exists, and realistic content where it does not. The important property is not the language but the length distribution: a design has to work with the shortest plausible entry and the longest one.
| Test with | Reveals |
|---|---|
| The shortest realistic text | empty space, broken alignment |
| The longest realistic text | truncation, overflow, wrapping |
| A single very long word | container overflow |
| No content at all | the empty state nobody designed |
The last row is the one most often skipped and most often shipped broken. Every list has a day when it is empty, and a placeholder-filled design never encounters it.
When is random text better than Latin?
When the content will be in a language with different characteristics, and when the text needs to be obviously fake. Random words in the target language give realistic line breaks and hyphenation, which Latin cannot.
It is also better where the placeholder must not be mistaken for content. Latin can look plausible in a document; obviously random text cannot, which makes it harder to ship by accident.
For anything with a character limit, neither matters as much as the length. Generating a placeholder of exactly the maximum permitted length is the test that finds the truncation.
Does it matter for accessibility?
Yes, in one specific way. Placeholder text left in a page is read aloud by a screen reader like any other content, so a page shipped with it produces a stream of meaningless Latin for anyone using one.
The same applies to placeholder images and to alt text that was never written. A design review that only looks catches the visible problems, and the ones that are only audible survive it.
It is another argument for real content early. Anything that has to be written eventually is cheaper to write before the layout is built around a stand-in.
Where does placeholder text still earn its place?
Early layout work, before any copy exists. A wireframe needs something with the texture of text, and asking for real words before the structure is settled produces copy that has to be thrown away.
It also has a role in testing systems rather than designs. Filling a database, checking how a component behaves with a thousand records, or exercising a search index all want text that is cheap to generate and obviously not real.
The line worth drawing is when the design starts being reviewed by anyone who will judge the content. From that point the placeholder is hiding the thing being reviewed.
Questions people ask
Why does it start mid-word? The passage was cut from the middle of a sentence, so the first surviving fragment is the tail of a longer word.
Are there themed alternatives? Plenty, and they are worse for the original purpose — a placeholder that is fun to read pulls attention back to the words.
Is the Latin meaningful? The original passage is. The scrambled version is not, and translating it is a dead end.
How much should I generate? Enough to fill the container and then somewhat more, since the overflow case is the one being tested.
Can I use it in a live prototype? For a click-through demo, yes. For anything a client might screenshot, use real text.
Does it affect search? A page indexed with placeholder text is indexed as a page about nothing. Remove it before publishing.
Fill the space, then test the extremes. The lorem ipsum generator produces the traditional Latin by paragraphs, sentences or words, and the random text generator produces the kind that cannot be mistaken for finished copy.