Editorial Policy

1. Who Writes the Content

Every article, tool description and transcription on this site is written by Eriberto Nascimento, a software developer with a long-standing interest in phonetics and second-language learning, and the founder of the site. There are no guest posts, no ghostwriters and no sponsored articles. When an article is updated, it is updated by the same person who wrote it.

We are not a university department and we do not claim to be. What we do claim is that the material is carefully sourced, that the limits of each tool are stated on the page, and that errors are fixed quickly when they are pointed out.

2. Where the Information Comes From

Phonetic claims on this site are checked against standard reference works rather than written from memory. The main sources are:

• Transcriptions: the Cambridge English Pronouncing Dictionary and the Longman Pronunciation Dictionary, which are the two reference works most widely used for British and American English.
• Symbols and articulation: the Handbook of the International Phonetic Association and the current IPA chart, published by the International Phonetic Association.
• Descriptions of English sounds: standard phonetics textbooks, including Ladefoged and Johnson's A Course in Phonetics and Roach's English Phonetics and Phonology.
• Teaching and learning claims: peer-reviewed work in second-language speech research, particularly on perceptual training, intelligibility and the difference between sounding native and being understood.

Where British and American pronunciations differ, both are given. Neither is presented as the correct one.

3. How Articles Are Produced and Reviewed

• Drafting. An article starts from a specific question a learner is likely to ask, not from a keyword. If we cannot explain why the answer matters for intelligibility, the article does not get written.
• Checking. Every transcription, minimal pair and stress pattern is verified against the reference works listed above before publication.
• Review. Articles are re-read on a rolling basis. The review date shown under the author's name on each post is the date the content was last checked, which is not always the date it was first published.
• Correction. When an error is found, the article is edited and the review date is changed. Corrections that change the substance of an explanation are noted in the article rather than made silently.

4. Use of AI Tools

We are direct about this because readers care. AI assistance is used for routine editorial work — tidying prose, checking consistency between articles, and building parts of the site's code. It is not used to invent phonetic facts, and no article is published without a human going through it and checking the claims against the sources named in section 2.

Responsibility for everything published here rests with the named author, whatever tools were used along the way.

5. Advertising and Editorial Independence

This site is free, requires no account, and is funded by advertising and affiliate commissions. The separation between that funding and the content works as follows:

• Advertisers have no influence over what is published. Ads are served programmatically by Google AdSense; we do not choose the advertisers and they do not choose our topics.
• No article on this site is paid for, sponsored or supplied by a third party.
• Ads are kept out of the reading flow in a way that would make the page harder to read, and are never disguised as article content, navigation or download buttons.
• Affiliate links — currently only to Preply, a tutoring marketplace — are disclosed in the text of any page that carries one, and are marked as sponsored links in the page code.
• A commission has never determined a recommendation. Where a free alternative is better for the learner, the free alternative is what the article says.

Details of the data collected by advertising partners, and how to opt out, are in our privacy policy.

6. Being Honest About What the Tools Can Do

Automatic pronunciation scoring has real limits, and we would rather say so than sell certainty we do not have. Browser speech recognition measures whether a recogniser produced the expected transcription — which correlates with good pronunciation but is not the same thing. It is weak on rhythm, intonation and fine vowel quality, and it can be thrown off by a noisy room or a distant microphone.

Several of the tools on this site deliberately tell you when a result is unreliable instead of showing a confident score, and the articles explain how to read a low score properly. No tool here will promise you a native accent.

7. Originality and Copyright

All text, illustrations and code on this site are original work produced for PronunciationChecker. Nothing is scraped, spun or republished from another site. Where we draw on published research or a reference dictionary, the source is named in the text rather than passed off as our own. If you believe any material here infringes your copyright, contact us and we will investigate and act promptly.