← Blog · 19 September 2026 · Lire en français

Reading everyday abbreviations aloud: etc., e.g., i.e., cf., NB

French prose is full of editorial abbreviations. Here is what a good text-to-speech engine should do with them, the mistakes to watch for, and the exact way to fix a one-off case.

A well-written French article carries more abbreviations than you would guess. Not units of measure, not organisation acronyms, but the everyday shorthand a writer drops in without thinking: "etc.", "cf.", "p. ex.", "c.-à-d.", "N.B.". Read on the page, they slip by unnoticed. Read aloud, each one is a decision: spell it out, expand it, or skip it? A text-to-speech engine that gets it wrong turns a careful sentence into gibberish, and that is exactly the kind of detail that gives away a generic voice tool the moment it meets real French copy.

Three families of abbreviation, three treatments

People often lump together three things that nothing forces you to read the same way. Acronyms and initialisms, like SNCF or NATO, are either spelled out or pronounced as words depending on custom. Unit and symbol abbreviations, like km/h or °C, expand according to the number in front of them. And then there is the third family, the quietest and the most poorly handled: editorial abbreviations, those Latin or French words shortened by typographic convention.

What sets them apart is that they are never read letter by letter. Nobody says "e-t-c" for "etc." The rule is easy to state and hard for a machine to keep: you pronounce the full word the abbreviation stands for, as if it had never been shortened.

The cases that keep coming up

Here are the most common forms in French copy, their expected reading, and the error an engine produces when it has no rule for them.

  • "etc." reads as et cetera. The classic mistake is to spell it, or worse, to treat the trailing dot as a full stop and cut the sentence dead.
  • "cf." reads as confer, or more naturally see. Spelled "c-f", it means nothing.
  • "p. ex." (the French "e.g.") reads as par exemple. The double dot trips engines up, which hear two separate abbreviations.
  • "c.-à-d." (the French "i.e.") reads as c'est-à-dire. The run of dots and hyphens is unreadable as-is unless a rule recognises it.
  • "N.B." reads as nota bene. Spelled out, it becomes a bare "n-b" the listener cannot place.
  • "P.-S." reads as post-scriptum, not "p-s".
  • "env." reads as environ (about), never "env".
  • "art." reads as article, which matters in legal and regulatory texts.
  • "éd." reads as édition or éditeur depending on context, and "vol.", "chap.", "p." become volume, chapitre, page.

Courtesy titles ("M.", "Mme", "Mlle") follow the same logic but raise their own questions, which we cover separately in our piece on courtesy titles in French TTS.

Why this is a real test for a tool

Two difficulties hide behind these tiny forms. The first is the dot. In written French, the period marks both the end of a sentence and an abbreviation. An engine that treats every dot as a full stop will break its breath in the middle of "p. ex." and leave an absurd gap. Reading it correctly means recognising that this particular dot is not a sentence boundary.

The second difficulty is context. A lone "p." can be page in a bibliography and nothing at all elsewhere. An honest engine does not guess: it expands what it recognises with confidence and leaves the rest exactly as written, rather than inventing. That caution, rather than cleverness, is what quality in French reading actually means: an abbreviation read as-is is better than an invented reading that misleads the listener.

What WeDispatch does, and how you stay in control

Everyday French abbreviations are expanded at read time, with no edit to your text. You write "c.-à-d." in your article the way you always have, and the listener hears "c'est-à-dire". The source file is never touched: the expansion happens at the moment of synthesis. This is the same discipline we apply to acronyms and initialisms and to units and symbols, each with its own rules.

That leaves the special case, the one no general rule covers: an in-house abbreviation, a usage specific to your field, a shorthand only you use. For that, the pronunciation lexicon lets you declare an entry: the written form on one side, the intended reading on the other. We walk through the full gesture in our article on fixing pronunciation with the lexicon. Once the entry exists, it applies to all your articles, present and future, with no per-article editing.

Hear how your own copy is read

The free plan lets you narrate your own articles and listen, on your real texts, to how abbreviations come out: try it now.

Give your articles a voice with WeDispatch

This blog is itself voiced by WeDispatch. Curious how it sounds on your content?

Book a demo

Read next