Emphasis (\emph)

When you want to stress part of a sentence, LaTeX gives you \emph{...}. It does not say “make this italic” — it marks the *meaning* “emphasize this.” By default that comes out italic, but used inside text that is already italic it switches the other way, back to upright (roman). This toggling is the heart of \emph, and it is exactly what sets it apart from \textit, which only ever slants.

Marking meaning, not appearance

Write \emph{the phrase to stress} and its argument is emphasized. The key point is that \emph specifies a role — “this is emphasized” — not a look such as “italic.” Which typeface stands for emphasis is decided by the document class or a package. The default happens to be italics, but that is just one part of how \emph is defined, not its essence.

This is exactly the distinction between <em> (emphasis) and <i> (italic) in web HTML. <em> carries the meaning “emphasized,” and how it actually looks is left to the CSS stylesheet. LaTeX works the same way: in the body you write the meaning with \emph, and you leave its appearance to the class file. That is the principle of separation of concerns — writing meaning apart from presentation — and it runs through the whole design of LaTeX.

The practical payoff is clear too. Because you marked the meaning, if you later decide that “emphasis should be bold, not italic,” you can change it everywhere by editing the single place that defines how emphasis looks — without touching the body at all. Sprinkle \textit{...} through the text instead, and every change of policy forces you to hunt down and rewrite every affected spot in the manuscript.

`\emph` toggles; `\textit` only slants

The most distinctive thing \emph does is toggle. When the surrounding text is ordinary upright (roman), \emph sets its argument in italics. But when the surroundings are already italic, that same \emph switches its argument back to upright. It automatically serves the goal “show emphasis in a way that stands out from the running text,” adapting to context: emphasis inside roman stands out as italic, emphasis inside italic stands out as upright — either way it stays visibly different from its surroundings.

By contrast, \textit{...} is a visual command that italicizes unconditionally. It always slants, never minding whether its surroundings are italic. So using \textit inside already-italic text changes nothing visually, and the word you meant to stress melts into the running text. This is the chief reason that **emphasis should use \emph, not \textit**. Reach for \textit when the look itself is the point — “this word is a book title, so set it in italics.”

Where \emph{...} is a command that takes an argument, \em performs the same emphasis as a declaration. You scope it with braces, as in {\em emphasized}. Like \emph, \em toggles between italic and upright according to context. For a short phrase the argument form \emph{...} is handier; for a long stretch the declaration \em is convenient. The pair \textit and \itshape is the same “command form versus declaration form” relationship.

Nested emphasis — it alternates

Sometimes you want to emphasize something inside an already-emphasized passage. Nest \emph, and its toggling nature makes italic and upright alternate at each level. The following example shows this.

\emph{start text \emph{middle text} end text}

Here the outer \emph sets start text and end text in italics. But the inner \emph{middle text}, sitting inside italic text, reverts to upright (roman), so middle text alone stands up straight. In other words the shape flips with nesting depth: “upright → italic → upright → ….” Nesting \textit does no such thing — however many layers you stack, everything stays italic and the inner emphasis is impossible to tell apart.

There is one more nicety: \emph inserts italic corrections automatically. When moving from a slanted glyph to an upright one, a tiny amount of space is added so the letters do not collide. Hand-placing \textit{...} next to upright text can crowd the glyphs unless you add that correction yourself with \/, but \emph understands its own boundaries and handles them. Simply marking the meaning brings this fine typographic care along for free.

Customizing the emphasis levels (`\DeclareEmphSequence`)

Alternating italic and upright at each nesting level is the default, but **\DeclareEmphSequence lets you design that freely. Introduced in the 2020 release of LaTeX, this relatively new mechanism takes a comma-separated list specifying which font declaration to use at each level of emphasis** (each depth of nesting). You put it in the preamble.

\DeclareEmphSequence{\itshape,%
  \upshape\scshape,\itshape}

In this example the first level of emphasis is italic, the second is small capitals, and the third is italic small capitals (provided you use a font that has those shapes). You could just as well build a hierarchy that fits your document — say, “first-level emphasis bold, the emphasis nested inside it italic.”

• If nesting goes deeper than the list provides — \emreset is used (by default \ulcshape\upshape, i.e. back to ordinary upright), and then the list restarts from the beginning.
• A “smart” check — LaTeX verifies that a given declaration actually changes the current font; if it would not, that level is skipped and the next is tried (on the assumption that the font was already changed there by hand).
• To override that check — add \emforce to the entry, which directs the mechanism to use it even when the font attributes appear unchanged.

Most documents need none of this — the default italic/upright alternation is plenty. What matters is that even when you do want to change the look, **the \emph in your body stays untouched**. The meaning of emphasis lives in the text; the way it is rendered lives in one place in the preamble — separation of concerns at work again. The full picture of commands that change the typeface itself is gathered on the “Font style commands” page.