Glossary

Phoneme

The smallest sound unit that changes meaning in a language. It is the alphabet used to write a pronunciation when spelling is not enough.

A phoneme is the smallest unit of sound capable of changing a word's meaning. In English, 'pat' and 'bat' differ by one phoneme, and that is enough to make them two words. French has around thirty-six of them, vowels and consonants combined, while its alphabet has only twenty-six letters, which is why spelling does not tell you the pronunciation.

Why this becomes concrete in a newsroom

Because proper nouns escape the rules. No reading system will guess how a French town name, a local surname or a regional brand is pronounced. It will apply the general rule and get it wrong, often on the word that matters most to the reader, the name of their own town.

The answer is not to respell the name phonetically in the article text. That fixes the sound by damaging the writing: the misspelled name would ship as-is into subtitles, transcript, sharing and the search index. The correction must live alongside the text, in a pronunciation lexicon where the newsroom records the expected pronunciation once, and the product finds it again afterwards.

How a pronunciation is written

Two notations coexist. The International Phonetic Alphabet is precise and unreadable to anyone who has not learned it. An approximate notation (writing roughly as it sounds) is immediately understandable by a sub-editor and is enough in the vast majority of cases. A tool aimed at a newsroom must accept the second; requiring IPA effectively reserves corrections to one person in the company, which guarantees they will not be made.

The real selection criterion

Not the richness of the notation, but PERSISTENCE. A correction that only applies to one article will have to be redone at every publication. The question to ask a vendor: 'when I fix the pronunciation of a name, does it apply to later articles without my thinking about it again?'

Related terms

  • Text normalization : The step that turns written text into speakable text: numbers become words, acronyms are expanded or spelled out, abbreviations are written in full.
  • Prosody : Everything in speech that is not the words themselves: rhythm, pauses, melody, intensity. It is what makes a reading sound human.
  • Text-to-speech (TTS) : Text-to-speech converts written text into spoken audio. Recent neural models produce voices that are hard to tell apart from human speech.

Give your articles a voice

WeDispatch automatically turns your articles into an audio version, the moment you publish. Try it free on your own articles: no credit card needed.

Book a demo