Editorial Policy & Sources

Sources and original analysis

Official Unicode and CLDR names, annotations, code points, versions, and categories are the factual starting point. Our conversational explanations are written independently from observed common usage, language conventions, platform context, and the ways a symbol can be misunderstood.

We do not copy descriptions from Emojipedia or vendor artwork. Product names such as WhatsApp, TikTok, Snapchat, Instagram, Discord, or iPhone are used only to explain where a symbol may appear or why a conversation can feel different.

Platform records distinguish an official app feature from cross-platform community slang, a cultural symbol, or a historical trend. They include a source, confidence label, and last-reviewed date; server-specific Discord customs and unsupported trend claims are not presented as universal meanings.

Local language and uncertainty

We prefer natural local wording and useful search phrases over literal translation. A locale marked as an editorial launch language has a prepared local seed set; a preview label means further native-language and cultural review is still needed. Preview pages remain noindex.

We describe likely readings with words such as usually, may, can, and depends. We do not infer romantic interest, gender intent, personality, or consent from one emoji. Relationship and surrounding words matter more than stereotypes.

Subtext meters and chat examples

Subtext meters use five named editorial bands, not percentages or measured probabilities. They are a baseline for common casual-chat usage in the selected language edition, and are never a psychological test, a prediction of someone’s intent, or evidence of attraction or consent.

Each published meter is paired with emoji-specific clear and risky chat examples. The words, relationship, region, generation, and platform can move the reading, so the examples explain the evidence instead of treating the sender’s gender as an answer. We do not show this feature in a language edition until its messages and explanations have been localized for that edition.

Updates and corrections

Content is reviewed when Unicode changes, platform rendering creates a meaningful difference, or readers provide a well-supported correction. Each mature emoji record should ultimately carry its own reviewer and review date rather than inheriting a site-wide build date.

Correction requests are evaluated for clarity, local usage, context, and possible regional differences. We may present more than one reading when communities use the same symbol differently.