There is no such thing as title case. There are four widely used title cases, they produce different results on the same headline, and every one of them is right inside the style guide that defines it.
The disagreement is narrow but it lands on words that appear in almost every headline, which is why two correctly capitalised versions of the same sentence can look like one of them is a mistake.
Where do the four disagree?
On short words, and almost nowhere else.
| Style | Articles | Conjunctions | Prepositions |
|---|---|---|---|
| AP | lowercase | lowercase if under 4 letters | lowercase if under 4 letters |
| Chicago | lowercase | lowercase if coordinating | lowercase regardless of length |
| APA | lowercase | lowercase if under 4 letters | lowercase if under 4 letters |
| MLA | lowercase | lowercase if coordinating | lowercase regardless of length |
The length rule is the fault line. AP and APA capitalise "Through" and "Between" because they run past three letters; Chicago and MLA lowercase them because they are prepositions and length is not the test. On a headline containing "Between", the two camps produce visibly different output.
All four agree on two things: the first word is always capitalised, and so is the last, whatever part of speech it happens to be.
What breaks title case?
Phrasal verbs, mostly. In "Log In to Your Account" the "In" is part of the verb rather than a preposition, so it takes a capital under every style — but no algorithm can tell that from the word alone, because "in" is a preposition in "A Walk in the Park" and a particle here.
The same problem hits "up", "out", "off" and "on". This is the reason automatic title casing needs a human read afterwards: a converter applies the word-list rule correctly and still gets "Set Up the Server" wrong if it treats "Up" as a preposition.
Proper nouns are the other reliable failure. Brand names that begin lowercase, or capitalise in the middle, all lose their intended shape when a rule capitalises the first letter of every major word.
Hyphenated compounds are the third argument. Chicago capitalises both halves of "Well-Known", AP capitalises the second element only when it is a proper noun or carries equal weight, and house styles routinely overrule both. Any converter has to pick one, so a hyphenated headline is worth reading back.
Why does uppercasing depend on language?
Because in Turkish the letter i has two capitals. Turkish distinguishes dotted and dotless i: uppercase "i" is "İ" and uppercase "ı" is "I", and getting it wrong changes the word.
This is not academic. Software that uppercases text without knowing the locale has produced real, shipped bugs — a string compared case-insensitively becomes a different string on a Turkish device, which is a well-documented class of failure in localisation work.
German has a similar corner. The sharp s, ß, has traditionally uppercased to "SS", so "Straße" becomes "STRASSE" — and since 2017 a capital ẞ is an officially permitted alternative, so both answers are defensible.
Numerals and units sit outside every rule. "3D", "iOS", "PDF" and "UK" are already in their correct form, and a title caser that lowercases the tail of an acronym produces "Pdf" — which is why a list of protected words is as important to a converter as the style rule itself.
What makes a good slug?
Lowercase, hyphens between words, ASCII only, and nothing else. A slug is a URL segment, so it inherits every constraint URLs have and adds the ones search results imply.
Diacritics get transliterated rather than dropped — "café" becomes "cafe", not "caf". Ampersands become "and" or vanish. Multiple spaces collapse to one hyphen, and hyphens do not stack. Leading and trailing hyphens are trimmed, which matters when the title ended in punctuation.
Underscores work technically and are worse practically: hyphens are the long-standing convention for separating words in a URL, and sticking to the convention costs nothing.
Is sentence case simply easier?
It is easier to apply and harder to get wrong, which is why most modern style guides for interface text prefer it. Only the first word and proper nouns take a capital, so none of the four-way disagreement above applies.
The catch is detecting where a sentence starts. Full stops appear in abbreviations and decimals, so converting a paragraph to sentence case mishandles "approx. 3.5 kg arrives Mon." in at least two places. Short strings are safe; long ones need checking.
Alternating case — capital, lowercase, capital — belongs to one internet joke and nothing else. It is worth knowing it exists so you recognise it, and worth knowing it renders as noise to a screen reader.
Questions people ask
Which style should I use? Whichever your publication already uses. Consistency inside one site matters more than the choice between them.
Does title case affect SEO? No. Search engines read the words, not the capitalisation. It affects whether a headline looks professional, which is a different and real thing.
Why does ALL CAPS read slower? Word shape disappears. Readers use ascenders and descenders to recognise words at a glance, and capitals remove that cue.
Are these conversions reversible? Lowercasing is lossy — the original capitals are gone. Convert from the original text rather than from a lowercased copy.
Pick the rule, then apply it. The case converter offers every mode at once, with uppercase, lowercase, title case and sentence case each isolated, the slug generator handles the URL rules including transliteration, and the mocking text generator does the joke.